Limiting Factors and Environmental Adaptability for Staple Crops in Kazakhstan

Abstract

Population growth increases the threat to global food security. Kazakhstan, a major agricultural nation, has made significant contributions to world food security; however, a wide gap exists between its food yield and that of other major crop-producing countries. Increasing food productivity in Kazakhstan through enhancing the utilization of natural endowments under existing cropland conditions would help alleviate global food pressure. Therefore, we elucidated the factors restricting Kazakhstan’s food productivity and proposed reasonable countermeasures. We analyzed the food production structure based on yearbooks. Correlation and stepwise regression were conducted on crop yield potential factors. The states of Kazakhstan were classified by hierarchical cluster and agronomic characteristics were evaluated using normalized scores. Wheat (60.3%), barley (14.9%), and potatoes (16%) are the main food crops produced in Kazakhstan. The ideal regional environment-based geographical crop configuration is “Northern—Wheat, Southern—Barley and Wheat, and Western—Potatoes.” The key limiting factors of wheat yield are water shortage and soil alkalization, while for barley, it is soil alkalization. The current planting distribution in Kazakhstan is suboptimal. Water-saving irrigation and agricultural runoff, staple crop planting layout optimization, organic fertilizer promotion, drought-resistant crop variety cultivation, and agricultural technology training must be prioritized to overcome crop yield constraints in Kazakhstan.

1. Introduction

Food security is essential for world peace and development and a foundation for building a community with a shared future for humankind. Ensuring food security has always been an enduring issue in the international community. After the conclusion of the MDGs (Millennium Development Goals) in 2015, the UN (United Nations) established SDGs (Sustainable Development Goals) to guide global development between 2015 and 2030 [1]. The subsequent goal of the SDGs was “zero hunger”, namely, the elimination of hunger, achievement of food security, improvement in nutrition, and promotion of sustainable agricultural development. However, the UN predicted that the global population will increase to nine billion by 2050 [2] and population growth will occur mainly in low-income and less developed countries. In general, these nations contend with more severe food security issues than developed countries [3]. The agricultural performance of major food-supplier countries has attracted considerable research attention to alleviate global food pressure [4].

Kazakhstan, known as the breadbasket of Central Asia and Eastern Europe, is a major food producer [5]. According to FAO (Food and Agriculture Organization of the United Nations) statistics [6], in terms of totals for 2021, cereal production is estimated at about 17 × 10⁶ t; among them, wheat is about 12 × 10⁶ t and barley is about 3 × 10⁶ t. Cereal exports in the 2021/22 marketing year (July/June) are forecast at about 7.5 × 10⁶ t; among them, wheat is about 6.5 × 10⁶ t and barley is about 80 × 10⁶ t. Due to high temperatures and insufficient rainfall in April and August, Kazakhstan’s cereal production and export volume dropped significantly in 2021, but Kazakhstan still makes a great contribution to world food security. However, Kazakhstan’s food productivity is not high, with a total cereal yield of only 1.37 t/ha, which is far from the traditional global impression of Kazakhstan (Figure 1). Before Kazakhstan gained its independence, it had undergone the “virgin land” movement. In 1956, 130 million hectares of wasteland was reclaimed to alleviate the food shortage in the Soviet Union [7].

Residual environmental and management problems, such as drought, water shortages, obsolete facilities, habitat loss, and soil degradation, created severe challenges for agricultural production during the early stages of independence in Kazakhstan. Implementation of various restorative measures made Kazakhstan’s grain production volatile over the past few decades. After experiencing environmental oppression and habitat degradation due to agricultural expansion, Kazakhstan’s Land Law and Environmental Law comprehensively stipulated the use of pesticides and the scope of arable land [8]. Further attention has been paid to the effective utilization and protection of land and water, the use of fewer fertilizers and pesticides, and biological (organic) plant protection methods to comprehensively control pests and erosion. Currently, Kazakhstan continues to implement extensive agricultural production management strategies [9]. However, inadequate irrigation capabilities and a fragile ecological environment remain an issue [10].

Exploring production potential by improving food productivity will effectively alleviate food pressure and curb environmental deterioration. Standard production includes potential research methods, such as the crop growth model, machine learning simulation [11], and yield gap theory (YGT). Among these, the YGT is more comprehensive and widely used. Research on agriculture in Kazakhstan is relatively limited, with many research gaps. Previous studies relied on remote sensing techniques and regional sample surveys [12,13] to examine water resource management, livestock production health, agricultural policy adjustment, and the Aral Sea ecology [14]. However, data collection and comprehension difficulties due to language barriers complicate the prospects of a comprehensive study of Kazakhstan.

In this study, we elucidated the limiting factors of food productivity in Kazakhstan and proposed reasonable countermeasures. Through the literature study, we summarized previous studies. We classified the 20 factors that may affect the YG into four categories: climate, soil, management, and socioeconomic, elucidating the limiting factors of food production in Kazakhstan through OLS and stepwise regressions. Considering the significance of food productivity on world food security and agricultural development in Kazakhstan, a new environmentally adaptive staple food production layout was constructed through hierarchical cluster analysis (HCA) and normalized score to avoid the environmental burden caused by expanding the land base. Our study aids in filling the gaps in the comprehensive agricultural research in Kazakhstan.

2. Food Production Status in Kazakhstan

2.1. Food Productivity Level in Kazakhstan

A research report ranking food production [15] by the leading global food-exporting countries indicated that Kazakhstan has meager crop yields but substantial growth potential [16]. This report prompted our research team to evaluate the contribution of Kazakhstan to worldwide food security. Based on the Knoema database [17], the top 15 cereal-exporting countries were selected as research objects, assuming that they represent global food suppliers (Figure 1). The production rates, yield, and harvest areas of major food crops were compiled in the FAOSTAT [6]. These include barley, buckwheat, maize, oat, potato, rice, rye, sorghum, millet, and wheat.

2.2. Food Production Structure in Kazakhstan

The predominant food crops in Kazakhstan between 2015 and 2019 were wheat (60.3%), corn (3.5%), barley (14.9%), rye (0.1%), oat (1.3%), buckwheat (0.3%), millet (0.2%), dried legumes (1.2%), rice (2.1%), and potato (16.0%) [18]. Wheat, barley, and potatoes are the staple crops and account for >90% of the total crop production in Kazakhstan. Low wheat yield reduces integrated food production, whereas the effects of barley and potatoes on this parameter are comparatively less evident. Wheat has a high planting ratio and there is substantial room to improve its yield; the overall food production capacity of Kazakhstan could be rapidly increased by enhancing wheat production (Figure 2).

2.3. Food Production Layout in Kazakhstan

The current production layout of Kazakhstan may be suboptimal. Wheat production is concentrated in Akmola, Kostanay, and North Kazakhstan, which comprise the major northern grain production base for the nation. Nevertheless, the states with comparatively higher wheat yields include Almaty, Zhambyl, and South Kazakhstan (renamed Turkestan in 2019). The states with the highest barley production are North Kazakhstan and Akmola. In terms of yield, however, Almaty, Zhambyl, East Kazakhstan, and South Kazakhstan in the southeast are the most conducive to barley cultivation. The main potato-producing states are Almaty, North Kazakhstan, and Pavlodar, but the states with the highest potato yield are Pavlodar, Karagandy, and Zhambyl (Figure 3).

3. Materials and Methods

3.1. Overview of Yield Gap

Potential yield, which refers to yield under the condition of good growth of crops of suitable varieties that are not subject to biotic and abiotic stresses, such as water, nutrients, pests, and diseases, and good management, is determined by the light and temperature conditions in a specific area [19]. YG is the difference between actual yield and potential yield. Studies on YG mainly focus on yield potential assessment, yield limiting factor analysis, and resource utilization efficiency between regions, which provide important information for continued food production growth and policy formulation.

The research on YG began in 1974 when IRRI (International Rice Research Institute, Los Baños, Philippines) studied the limiting factors of Asian rice. Gomez [20], one of its Research members, proposed the preliminary concept of YG in 1977. In 1981, de Datta [21] defined YG as the gap between the actual yield in farmland (AYF) and the potential yield in experimental stations (PYES), which was divided into two levels: the gap between the PYES and the potential yield in farmland (PYF, technology and environment that cannot be applied and realized in farmland); the gap between potential and actual farmland yields, as caused by biological constraints (variety, weeds, diseases and insects, soil problems and fertility, water) and socio-economic constraints (cost and returns, credit, tradition and attitudes, knowledge, input availability, institutions). In 1984, Fresco [22] introduced the concept of “economic upper bound yield” and extended research on YG to the scope of agricultural systems. In 2000, de Bie [23] proposed the upper limit of “potential yield of simulated experimental station” and YG was divided into three levels: simulated YG, experimental station YG, and farmland YG. In 2009, Lobell [24] summarized the quantification of yield potential: model-simulated yield, high-yield record, experimental yield, and higher farmer yield. In 2010, Fischer [25] et al. proposed “developable YG”, adding the concept of “attainable yield”.

Research on the limiting factors of crop yield tends to be concentrated in the field of agronomy, which can be summarized as a discussion of environmental factors (focus on climate and soil) and planting management. In a study on the sugarcane YG in tropical islands, years of climate and soil data were used to calculate crop growth simulations [26]. In analyzing field management factors, germplasm, plant density, fertilizer, hybridization, irrigation [27], and sowing time are all potential factors that can affect productivity. A Colombian cattle breeding system study used climatic, edaphic, and land characteristics [28] to determine agro-ecological zones (AEZ). Previous studies on agricultural practice integration (API) to reduce the rice yield gap in Africa have focused on quantifying the impact of crop variety, crop establishment, nutrients, weeds, and water on yield [29]. Rong et al. summarized the methods for quantifying the yield gap worldwide for wheat, maize, and rice and argued that climate, nutrients, moisture, variety, planting date, and socioeconomic factors were the most frequently mentioned [30].

Although the vast majority of studies on the yield gap focus on the field of agronomy, the socioeconomic field is also an important part of crop yield. Another study on the maize YG in Ethiopia focused on market imperfections, economic constraints, and management policies [31]. The recently developed Shared Socioeconomic Pathways (SSP) framework fully considers population, economy, technology, and environment [32] and distinguishes five socioeconomic statuses: sustainability development (SSP1), middle of the road trends (SSP2), fragmentation (SSP3), inequality (SSP4), and conventional development (SSP5). Wu et al. summarized the factors influencing poor winter WY in the world as: (1) varieties, such as drought-resistant varieties and varieties with high nitrogen utilization efficiency, (2) climate, which causes inter-annual or inter-regional yield gaps, (3) soil, including thickness of plow layer, water content, nutrients, and organic matter content, (4) management measures, such as planting systems, plant protection management, fertilization, and irrigation, and (5) socio-economic and political factors, landscape, ecosystem, and cultivated area [33].

In summary, the influencing factors of yield gap were divided into four categories (Table 1): climate (precipitation, temperature, wind speed, radiation, snow cover), soil (salinization, alkalization, wind and water erosion, fragmentation), management (water and fertilizer) and socioeconomics (market profitability, labor, infrastructure) in this study.

3.2. Methods

3.2.1. Analysis of Influencing Factors

Correlation analysis of possible yield impact factors was performed through:

$$r=\frac{\sum (x-\overline{x})\left(y-\overline{y}\right)}{\sqrt{\sum {\left(x-\overline{x}\right)}^{\mathbf{2}}·\sum {\left(y-\overline{y}\right)}^{\mathbf{2}}}},$$

(1)

where r is the correlation coefficient between factors X and Y. The factors that strongly correlated with WY, BY, and PY were selected for regression analysis. Multiple regression (based on ordinary least squares) analysis was implemented by [34]:

$${\mathrm{Y}}_{\mathrm{i}}={\mathsf{\beta }}_0+{\mathsf{\beta }}_1{\mathrm{X}}_{1\mathrm{i}}+{\mathsf{\beta }}_2{\mathrm{X}}_{2\mathrm{i}}+\cdots +{\mathsf{\beta }}_{\mathrm{k}}{\mathrm{X}}_{\mathrm{ki}}+{\mathsf{\mu }}_{\mathrm{i}},$$

(2)

where k is the number of explanatory variables, i is the number of sample groups, ${\mathsf{\beta }}_0$ is the constant term, and ${\mathsf{\mu }}_{\mathrm{i}}$ is the residual term. The insignificant explanatory variables were removed using stepwise regression.

3.2.2. Analysis of Environmental Adaptability

According to the remaining significant impact explanatory variables, the hierarchical clustering [35] of 14 states was performed by:

$$\mathrm{D}=\sqrt{{\left({\mathrm{M}}_1-{\mathrm{N}}_1\right)}^2+{\left({\mathrm{M}}_2-{\mathrm{N}}_2\right)}^2+\cdots +{\left({\mathrm{M}}_{\mathrm{k}}-{\mathrm{N}}_{\mathrm{k}}\right)}^2},$$

(3)

where D is the Euclidean distance between samples M and N, and k is the number of significant impact explanatory variables. For the clustering results, 14 states were grouped and each significant explanatory variable was normalized by

$$\mathrm{S}=\frac{{\mathrm{K}}_{\mathrm{i}}-\min \left({\mathrm{K}}_{\mathrm{i}}\right)}{\max \left({\mathrm{K}}_{\mathrm{i}}\right)-\min \left({\mathrm{K}}_{\mathrm{i}}\right)},$$

(4)

where $\mathrm{S}$ is the score of a state on the significant explanatory variable K. The scores of the 14 states in each factor are represented by a composite radar chart.

“RStudio” (https://www.rstudio.com/products/rstudio/download/ (accessed on 20 June 2021)) and “Excel” (Microsoft Corp., Redmond, WA, USA) were used for all data calculations, while “ArcGIS” (Esri, Redlands, CA, USA) and “Origin” (OriginLab, Northampton, MA, USA) were used to plot the drawings.

3.3. Data Sources

Data were sourced from the Kazakhstan Agricultural and Environmental Statistical Yearbook from 2010 to 2019, titled Agriculture, Forestry and Fisheries in the Republic of Kazakhstan [18] and Environmental Protection and Sustainable Development of Kazakhstan [36], respectively. The specific secondary indicators, abbreviations, and data preprocessing of 20 factors are shown in Table 1.

Table 1. Impact factors of Kazakhstan’s yield.

Cxategories	Factors	Abbreviations (Unit)	Data Processing	Descriptive Statistics
Climatic factors [37]	Annual total precipitation	ATP (mm)	Raw data from yearbooks.	Mean: 309.76; SD: 136.76; Min: 78; Max: 984.
	Annual average temperature	AAT (°C)		Mean: 7.06; SD: 3.62; Min: 0.70; Max: 13.60.
	Annual average wind speed	AWS (m/s)		Mean: 2.81; SD: 1.01; Min: 0.00; Max: 4.40.
	Annual total radiation	ATR (106 cal/cm2)		Mean: 135.41; SD: 32.10; Min: 80.40; Max: 224.83.
	Annual snow cover	ASC (cm)		Mean: 25.2; SD: 13.42; Min: 1; Max: 55.
Soil health factors [38,39]	Land salinization	LSA (%)	Proportions of salinized, alkalized, wind-eroded, and water-eroded (excessive humidity, swamp, and washout) areas in agricultural land.	Mean: 18.42; SD: 15.54; Min: 5.52; Max: 58.13.
	Land alkalization	LAA (%)		Mean: 25.27; SD: 15.19; Min: 3.45; Max: 55.67.
	Wind erosion	WIE (%)		Mean: 13.53; SD: 11.78; Min: 0.00; Max: 34.39.
	Water erosion	WAE (%)		Mean: 4.40; SD: 2.53; Min: 0.42; Max: 10.64.
	Fragmentation	FA (%)		Mean: 16.98; SD: 15.75; Min: 1.40; Max: 53.53.
Management factors	Agricultural irrigation water	IWA (m3/ha);	Irrigation water (IW) divided by crop area;	Mean: 2785.04; SD: 5381.74; Min: 0.94; Max: 23,511.74.
	Mineral fertilizer application ratio	MFR (%)	Ratio of mineral fertilizer and organic fertilizer application area to crop area, respectively.	Mean: 6.42; SD: 6.80; Min: 0.09; Max: 27.22.
	Organic compound application ratio	OCR (%)		Mean: 0.11; SD: 0.18; Min: 0.00; Max: 1.04.
	Mineral fertilizer application intensity	MFI (t/ha)	Raw data from yearbooks.	Mean: 171.67; SD: 249.90; Min: 3.10; Max: 1611.80.
	Organic compound application intensity	OCI (t/ha)		Mean: 16.24; SD: 22.85; Min: 0.10; Max: 144.40.
Socioeconomic factors [40]	Rural water supply services	WSR (%)	Raw data from yearbooks.	Mean: 87.88; SD: 5.75; Min: 75.70; Max: 100.00.
	Electricity production	EP (kW·h/ha)	Power generation (PG) divided by state area.	Mean: 396.71; SD: 759.70; Min: 34.60; Max: 3657.30.
	Agricultural economic practitioners	AEP (company/103 ha)	Numbers of agricultural economic activity participants and offices (AEPO) and farmers (TF) divided by crop area.	Mean: 7.30; SD: 25.54; Min: 0.09; Max: 160.00.
	Farmers	F (company/103 ha)		Mean: 136.34; SD: 418.43; Min: 0.59; Max: 2660.00.
	Crop production profit margin	CPA (%)	Raw data from yearbooks.	Mean: 29.40; SD: 20.81; Min: −21.70; Max: 80.90.

Table 1. Impact factors of Kazakhstan’s yield.

Cxategories	Factors	Abbreviations (Unit)	Data Processing	Descriptive Statistics
Climatic factors [37]	Annual total precipitation	ATP (mm)	Raw data from yearbooks.	Mean: 309.76; SD: 136.76; Min: 78; Max: 984.
	Annual average temperature	AAT (°C)		Mean: 7.06; SD: 3.62; Min: 0.70; Max: 13.60.
	Annual average wind speed	AWS (m/s)		Mean: 2.81; SD: 1.01; Min: 0.00; Max: 4.40.
	Annual total radiation	ATR (106 cal/cm2)		Mean: 135.41; SD: 32.10; Min: 80.40; Max: 224.83.
	Annual snow cover	ASC (cm)		Mean: 25.2; SD: 13.42; Min: 1; Max: 55.
Soil health factors [38,39]	Land salinization	LSA (%)	Proportions of salinized, alkalized, wind-eroded, and water-eroded (excessive humidity, swamp, and washout) areas in agricultural land.	Mean: 18.42; SD: 15.54; Min: 5.52; Max: 58.13.
	Land alkalization	LAA (%)		Mean: 25.27; SD: 15.19; Min: 3.45; Max: 55.67.
	Wind erosion	WIE (%)		Mean: 13.53; SD: 11.78; Min: 0.00; Max: 34.39.
	Water erosion	WAE (%)		Mean: 4.40; SD: 2.53; Min: 0.42; Max: 10.64.
	Fragmentation	FA (%)		Mean: 16.98; SD: 15.75; Min: 1.40; Max: 53.53.
Management factors	Agricultural irrigation water	IWA (m3/ha);	Irrigation water (IW) divided by crop area;	Mean: 2785.04; SD: 5381.74; Min: 0.94; Max: 23,511.74.
	Mineral fertilizer application ratio	MFR (%)	Ratio of mineral fertilizer and organic fertilizer application area to crop area, respectively.	Mean: 6.42; SD: 6.80; Min: 0.09; Max: 27.22.
	Organic compound application ratio	OCR (%)		Mean: 0.11; SD: 0.18; Min: 0.00; Max: 1.04.
	Mineral fertilizer application intensity	MFI (t/ha)	Raw data from yearbooks.	Mean: 171.67; SD: 249.90; Min: 3.10; Max: 1611.80.
	Organic compound application intensity	OCI (t/ha)		Mean: 16.24; SD: 22.85; Min: 0.10; Max: 144.40.
Socioeconomic factors [40]	Rural water supply services	WSR (%)	Raw data from yearbooks.	Mean: 87.88; SD: 5.75; Min: 75.70; Max: 100.00.
	Electricity production	EP (kW·h/ha)	Power generation (PG) divided by state area.	Mean: 396.71; SD: 759.70; Min: 34.60; Max: 3657.30.
	Agricultural economic practitioners	AEP (company/103 ha)	Numbers of agricultural economic activity participants and offices (AEPO) and farmers (TF) divided by crop area.	Mean: 7.30; SD: 25.54; Min: 0.09; Max: 160.00.
	Farmers	F (company/103 ha)		Mean: 136.34; SD: 418.43; Min: 0.59; Max: 2660.00.
	Crop production profit margin	CPA (%)	Raw data from yearbooks.	Mean: 29.40; SD: 20.81; Min: −21.70; Max: 80.90.

4. Results

4.1. Influencing Factors of Food Productivity in Kazakhstan

Factor correlation analyses revealed certain phenomena (Figure 4). (1) There were obvious correlations among climate (annual total precipitation (ATP), annual average temperature (AAT), annual average wind speed (AWS), annual total radiation, and annual snow cover) and soil health (LSA, LAA, WIE, WAE, and FA) factors. Soil salinization (LSA) was positively correlated with AAT but negatively correlated with annual snow cover and ATP. (2) Among the management condition factors (IWA, MFC, OCC, MFA, and OCA), only irrigation (IWA) was strongly correlated with soil health and climate factors. (3) Of the market factors (agricultural economic practitioners (AEP), farmers (F), and CPA), F and AEP (agribusiness) were strongly correlated with the climate and soil health factors. (4) In crop productivity, WY was strongly correlated with the climate and soil health factors. The results of the correlation analysis of BY and WY were similar. However, PY was positively correlated with power generation (EP) and soil fragmentation (FA). Hence, the subsequent regression analysis selected ATP, AAT, AWS, LSA, LAA, WIE, WAE, FA, IWA, EP, AEP, F, and CPA.

Regression coefficients accurately reflect the influences of various factors (Figure 5). (1) Among the weather factors, increases in ATP strongly promoted wheat, barley, and potato yield, AAT promoted wheat yield, and AWS promoted potato yield. (2) Poor soil health inhibited crop yield. LSA severely inhibited wheat yield and soil salinization (LSA), soil alkalization (LAA) severely inhibited wheat and barley yield, and water erosion (WAE) inhibited barley and potato yield. FA inhibited wheat and barley yield but promoted potato yield. (3) In terms of the management factors, increases in agricultural irrigation water (IWA) promoted wheat yield but not barley or potato yield and an adequate power supply promoted potato yield. (4) Concerning the market factors, AEP and farmers (F) had a variable impact on crop yield (both are unstable in OLS and AEP has negative effects in stepwise model), whereas increases in the crop production profit margin for agricultural enterprises (CPA) promoted crop yield.

To sum up, the following rules can be concluded: (1) LSA and LAA have the greatest impact on wheat, followed by barley, and potatoes were the most resistant to harsh environmental conditions; (2) wheat is the most water intensive, followed by barley, and potatoes the most drought-tolerant. Furthermore, WIE was nonsignificant in the OLS and ineffective in stepwise regressions. Thus, it was omitted from the classification reference (Figure 5). At the same time, the results of stepwise regression are more precise, which is beneficial to combine with the standardized score results.

4.2. Environmental Adaptability of Staple Crops in Kazakhstan

Combined with the cluster analysis results, we divided the 14 states into five categories and the scores are shown in Figure 6.

Mangystau, Kyzylorda, and Atyrau are the poorest agricultural environments as they have low ATP, high AAT, and severe LSA. Based on the agricultural distribution map of Kazakhstan (Figure 3), the high wheat yield in Kyzylorda was achieved at the expense of IWA (more detailed explanations in the Discussion section), which should not be advocated.

The environment of Kyzylorda is comparatively more suitable for growing potatoes and barley. Potatoes are comparatively less sensitive to drought and LSA. Kyzylorda has milder LAA than most other states, so it cultivates soil alkalization-sensitive barley. Atyrau is appropriate for planting potatoes as this crop is the least sensitive to LAA. Mangystau is unsuitable for growing any previous crops due to severe water shortages. Nevertheless, many agricultural workers reside in this state (AEP, F).

Aktobe, Karagandy, and West Kazakhstan are located along the horizontal midline of Kazakhstan. They have lower precipitation rates (ATP), severe LAA, mild LSA, and limited available IWA. Therefore, these states are suitable for growing potatoes.

Akmola, Kostanay, and North Kazakhstan constitute the national breadbasket in northern Kazakhstan. They have moderate annual precipitation (ATP), low AAT, slight LSA, severe LAA, springtime freeze–thaw (WAE), insufficient power (EP), and no reliance upon irrigation (IWA). Therefore, these states are more conducive to cultivating wheat than barley and potatoes, because wheat consumes the least irrigation water here.

Almaty, Turkestan, and Zhambyl comprise the national fruit and vegetable production center in southern Kazakhstan. Their agricultural environment is fit for all crop production. They have abundant rainfall (ATP), high average annual temperature (AAT), comparatively normal LSA and LAA, severe WAE, and a lack of electricity (EP). They are also the three states with the highest barley and wheat yields (Figure 3). Thus, they are appropriate for wheat and barley cultivation.

Pavlodar has severe LAA but moderate water (ATP) and power (EP) for crop cultivation; therefore, it is suitable for planting potatoes. All parameters, except FA, are moderate in East Kazakhstan. For this reason, this state is fit to produce all three crops.

In conclusion, the “Northern Wheat, Southern Barley and Wheat, Western Potatoes” crop planting distribution is appropriate for the natural environment of Kazakhstan.

5. Discussion

5.1. Model Evaluation

The stepwise model effectively eliminates insignificant and severe collinearity factors (

Table 2. VIF values of regression models.

Factor	OLS/WY	Stepwise/WY	OLS/BY	Stepwise/BY	OLS/PY	Stepwise/PY
ATP	2.893	2.057	2.979	2.273	2.712	2.537
AAT	6.488	2.127	7.453	/	6.583	/
AWS	4.097	/	4.688	/	4.325	3.865
LSA	16.042	4.534	14.671	/	11.533	/
LAA	10.303	5.151	9.929	4.66	9.673	4.516
WIE	12.906	/	7.224	/	7.112	/
WAE	4.337	/	4.371	3.382	4.295	3.937
FA	8.487	3.071	8.76	3.041	8.718	6.013
IWA	16.98	/	8.419	4.172	9.17	5.968
EP	2.454	1.221	1.791	1.27	1.825	1.327
AEP	25.168	1.365	32.141	1.766	7.13	7.071
F	37.237	/	38	/	7.922	7.655
CPA	1.821	1.701	2.22	1.873	1.95	1.819

). VIF (variance inflation factor) is a measure of the severity of multicollinearity in multiple linear regression models. In this study, the tolerance is 0.1, so VIF is the reciprocal of tolerance, which is 10. All OLS models have VIF values over 10; however, after removing factors through the stepwise model, most of the VIF values are below 5 (almost no collinearity) and only a few are between 5 and 10 (collinearity is not serious). Meanwhile, after the stepwise model treatment, all the remaining factors are significant.

After repeated verification of factor elimination, it was found that there was negative collinearity between LSA and IWA. When one of the LSA or IWA factors was removed randomly, the influence of the other variable is reduced and the VIF decreased substantially. This shows that increasing irrigation can alleviate the impact of soil salinization. However, this method is not desirable, because massive irrigation itself will lead to the loss of soil nutrients, resulting in salinization.

Here, certain independent variables were replaced to evaluate the stability of the regression models. The original indices reported in the yearbooks, including agricultural economic activity participants and offices (AEPO), total power generation in different regions (PG), and total number of farmers per state (TF), were imported into the stepwise models for comparison. AEPO and AEP have opposite effects on BY and PY, and the impact measurement of market indicators is unstable. Therefore, this study avoided excessive analysis of market indicators in Section 4.2.

5.2. Results Analysis and Related Research

Interesting phenomena appeared in the regression results, such as the promotion of potato yield by FA, inhibition of barley and potato by IWA, and inhibition of wheat and potato by AEP. Soil health is vital to agricultural production and regeneration; soil degradation and erosion must be avoided or corrected to the furthest extent possible [41]. However, the climate in Kazakhstan is drastic and an area may experience drought and water erosion at near periods [36]. After index analysis, it can be found that potatoes are resistant to drought but not excessive water. There must be a particular offset between the drought reflected by FA and ATP and the excessive water reflected by WAE and IWA and further path analysis will better explain this problem. There are two reasons for the large impact deviation of AEP (company/10³ ha). One is that there are agricultural practitioners in Mangystau and Atyrau states but almost no food production activities; the second is that agricultural firms are disproportionately concentrated in Turkestan (22%), but its wheat and potato yields are not much better than those of other states.

In recent years, research on food production in Kazakhstan has been minimal. A model designed to assess the impact of climate change on spring wheat production in Kazakhstan and Russia disclosed that lack of water is the main factor restricting yield during the growing season [42]. Another study on the effects of climate change on barley and wheat yield in Kazakhstan revealed that extreme heat events had minimal impact on these parameters. However, the combination of heat and drought reduced yield [43]. The authors recommended planting drought-tolerant crops in areas where climate change has had a negative impact on yield. Nevertheless, certain research groups suggested that the grain production base in northern Kazakhstan is stable and slight increases in rainfall will be offset by increased evaporation and have a low net impact on yield [44]. Previous studies focused on remote sensing and meteorological data in Central Asia. Language barriers have impeded comprehensive studies on Kazakhstan. The present study was a systematic macro-level assessment of the overall food production capacity of Kazakhstan. (1) Three representative crops were considered, rather than the single-crop approach used in previous studies. (2) Consideration was given to climate, soil, management, infrastructure, and market. (3) The novel layout “Northern Wheat, Southern Barley and Wheat, Western Potatoes” was proposed and analyzed from an environmental perspective.

5.3. Countermeasures and Suggestions

(1)

Agricultural technology development and introduction in arid areas.

Though wheat accounts for 60.3% of the food output in Kazakhstan, its yield is only 0.99 t/ha, which falls far below the international level [6]. The Global Agroecological Zones model projected that wheat productivity could be increased in this region to ≥70.3% [16]. Regression analysis showed that water shortage is a major limiting factor in wheat production in the area. Precipitation (ATP) and IWA positively affect wheat yield.

This nation is one of the driest on the Eurasian continent and it borders numerous regions with water management issues. However, agriculture accounts for ≤70% of the total water consumption in all sectors [45]. Although ATP is relatively low in Kazakhstan, it is concentrated from April to June [46]. Furthermore, melting snow in the springtime causes water levels to rise excessively in rivers and lakes [47]. The foregoing water inputs frequently lead to soil erosion in northern and eastern Kazakhstan. Therefore, reservoirs and pipelines should be installed in lakes and areas with concentrated precipitation.

The agricultural ecology of the west is highly fragile and challenged by Aral Sea desertification, dust storms, and soil degradation caused by early-season over-cultivation. Therefore, drip irrigation technology with maximum water-saving efficiency is recommended for the arid zones in the west. However, film mulching technology is not recommended for this region, contributing to agricultural pollution. The challenge in adopting degradable mulches is that the currently available mulches cannot simultaneously provide a balance between certified materials for use in organic production and materials that allow desirable agricultural yields [48].

(2)

Adjustment of crop production distribution.

Studies on the distribution of staple crops in Kazakhstan showed that the current planting management does not optimally use the production environment. Regression analyses demonstrated that wheat, barley, and potato require entirely different growing conditions. Transportation, storage, and labor factors might contribute to the current unsuitable planting distribution in Kazakhstan. “Northern Wheat, Southern Barley and Wheat, Western Potatoes” is also consistent with the results of another study on the impact of climatic factors on three staple crops in Kazakhstan [49]. Potato is suitable for cool temperatures but not excess water. In Kazakhstan, high latitudes are cold and abundant in ATP, while low latitudes are hot and summer precipitation is concentrated. Only the mid-latitude regions meet the plant physiological properties of potatoes. In addition, high latitudes are too cold to grow winter barley, while low latitudes meet the growing conditions of winter barley and spring barley.

The following detailed distribution may be consistent with the planting environment in Kazakhstan. (1) Kostanay, North Kazakhstan, and Akmola should be sown with wheat. (2) Turkestan, Zhambyl and Almaty should cultivate barley and wheat. (3) Kyzylorda should grow barley and potato, as this state is arid and wheat has a higher water demand. (4) Pavodar and Karagandy are suitable for potato. (5) Aktobe, West Kazakhstan, and Atyrau should plant potatoes as this crop has a low ecological footprint and is the most resistant to harsh environmental conditions. These states are close to the Aral Sea and have serious soil degradation [10]. (6) East Kazakhstan makes flexible choices according to people’s food needs.

(3)

Organic fertilizer and drought-resistant crop varieties should be promoted.

In Kyzylorda, there is little rainfall, but enormous volumes of IWA are applied to maximize wheat yield there. This management practice must cease or else it will result in secondary soil salinization. Hence, water conservancy must be improved in Kyzylorda.

Long-term chemical fertilizer application may leave saline and alkaline residues to compact the soil. By contrast, the humic substances in organic fertilizers adjust the pH, activate the microorganisms, and improve the physicochemistry in the soil. Therefore, using organic fertilizers can alleviate soil salinization caused by excessive irrigation and overgrazing in western and central Kazakhstan, respectively [50]. Mineral soil amendments with rapid onset may also be implemented at the initial cultivation stages.

Researchers and growers worldwide acknowledge that water is the single most crucial abiotic factor limiting crop productivity in the face of global climate change. Biotechnology is promising for developing novel drought-resistant crop varieties that could help foster climate-secure agriculture [51]. As agricultural demonstration parks have been established abroad, numerous high-quality germplasm resources have been screened and exchanged. It is expected that field trials will be conducted to develop cultivars amenable to the arid zones in Kazakhstan.

(4)

Agricultural technology training and service promotion.

The analysis of the influences of F and AEP density is not stable. Nevertheless, the professional skills of agricultural workers must be upgraded. Even with the large-scale operation of agricultural companies [52], the idle rate of arable land in Kazakhstan is still high, which means the existing management efficiency of farmers is not enough to cover the amount of arable land resources [36]. Farmers are introduced to new agricultural technologies via training and practical application. Promoting modern technologies through education expands farm management capacity and empowers farmers. Furthermore, trained farmers can disseminate and promote the adoption of novel agricultural service systems among their colleagues via social networks and education programs [53]. Online and offline guidance should be available for weather forecasting, planting and grazing cycle control, fertilization management, and cultivation methods conducive to the development of local green agriculture.

6. Conclusions

Wheat, barley, and potato are the staple crops produced by Kazakhstan. The key limiting factors of wheat yield are water shortage and soil alkalization, while for barley, the main issue is soil alkalization. Although the current geographical distribution of staple crops is suboptimal in Kazakhstan, the proposed “northern—wheat, southern—barley and wheat, and western–potatoes” distribution is conducive to high crop productivity and is based on the climatological and pedological properties of each state and provides supporting suggestions. Governments can reduce the waste of agricultural resources owing to socio-economic development by coordinating infrastructure and state agricultural planning directions. The combination of HCA and normalized scores to determine the environmental adaptability of states can also provide macro-research ideas for other scholars. However, this study has some data limitations; the indicators of yearbook statistics are extensive and non-selective (time and content). Research on the extensive causes of the concerned phenomena in the regression results requires detailed information on the soil physicochemistry, air quality, and farming methods [54,55]. As the Northwest A&F University endeavors to cooperate with other agricultural universities in Kazakhstan and agricultural demonstration gardens are well established overseas, additional systematic field trials are anticipated in the future.