Sustainability is commonly described as resting upon three pillars, i.e., social, economic, and environmental concepts (Purvis et al., 2019). Thus, equal consideration should be given to social equity, livability, community development, etc. (social sustainability) as to long-term economic growth and development (economic sustainability), and carbon footprint and biodiversity (environmental sustainability). However, this approach has been criticized for dividing sustainability into separate realms, and the necessity to find an integrated way forward has been highlighted (Boyer et al., 2016). In an agricultural context, sustainable crop production systems involve enhanced biodiversity and living soils, use of integrated pest management, and low greenhouse gas emissions, while simultaneously producing high-quality food and sustaining food security (Imadi et al., 2016). Sustainable and successful crop production should thus comprise: (i) development of farming and stable livelihoods for farmers, together with (ii) high biodiversity and low carbon footprint, and (iii) high, secure, and safe yield of high quality and stability in provision of food for end-users. Various factors can negatively impact the security, safety, and quality of crop production, with diseases, pests, weeds, animal pests, and abiotic factors being the most important factors (Oerke, 2006; Savary et al., 2012). Strategies to combat these biotic and abiotic stresses, without affecting biodiversity, soil micro-environments, and carbon footprint, are essential. Continuous monitoring and strategy development is required for managing different agricultural crops at national and international levels (Park et al., 2011; Sikharulidze et al., 2015; Bouwmeester et al., 2016).

Tajikistan is amongst the poorest countries in the world, with high levels of hunger and undernourishment (FAO, 2015; WFP, 2017). Wheat is the major staple food crop in Tajikistan, contributing 60% of the daily calorie intake of the Tajik people (Braun, 2003). The crop is widely grown throughout Tajikistan, from 300 m above sea level (masl) in Tajik Delvarzin in the north-west, near the border with Uzbekistan, to 3,000 masl in Western Pamirs (Muminjanov et al., 2016). According to the latest official figures from the State Agency of Statistics of Tajikistan, wheat was grown on 276 000 ha in 2018, representing 70% of all areas under cereals and pulses. Overall wheat production in 2018 was 779 000 tons, with an average yield of about 2.8 t/ha (TajStat, 2018). Despite the clear dominance of wheat over other crops in Tajikistan, domestic production cannot meet domestic demand, so imports, primarily from Kazakhstan, are required. Price fluctuations and instability on the world market led to increased domestic wheat production being set as a major strategic goal by the Tajik government. However, the current wheat yield in Tajikistan is low in comparison with that in neighboring countries, e.g., in Uzbekistan the average yield is 4.7 t/ha (FAO-REU, 2016).

Tajikistan is known as a hotspot of biodiversity due to its varied mountainous landscape, with habitats ranging from deserts to glaciers and frozen mountains (Fauna Flora International, 2019). The Pamir regions are known as a rich habitat for local tree species and varieties of e.g., apple, apricot, mulberry etc. (Giuliani et al., 2011). Tajikistan is also the home of many unique wheat types, including dwarf wheat and sphere grain wheat, and of most legumes (peas, lentil, chickpea, common bean, golden bean, horse bean), and is the center of diversity for rye, mustard, flax, safflower, cotton-guza, etc. (Muminjanov, 2008). However, Tajikistan is also known as a “cradle” of weed flora, i.e., as the center of origin, and is thereby rich in endemic plants (Nowak et al., 2011, 2013). This widespread presence of weeds is a serious issue for wheat production in Tajikistan, hampering safe and secure food production for the population through decreased yield and through the harvested grain comprising a mixture of wheat and weed seeds. Weed management is currently completely lacking in many wheat fields and, when carried out at all, it involves either hand weeding or hoeing (Nowak et al., 2013).

Another issue for crop production in Tajikistan is the negative impact of pests and diseases on yields. Wheat rusts, especially stripe rust (Puccinia striiformis f.sp. tritici) and leaf rust (P. recondite f.sp. tritici), are the major diseases negatively affecting the wheat crop in Tajikistan, followed by powdery mildew (Blumeria graminis f. sp. tritici), tan spot (Pyrenophora tritici-repentis as a complete stage), and seed-borne diseases such as common bunt (Tilletia laevis; T.tritici) and loose smut (Ustilago tritici) (Pett and Muminjanov, 2006a; Pett et al., 2006). Significant yield losses due to stripe rust outbreaks have been reported during epidemic years (Rahmatov et al., 2010). Further, yield and quality reductions in wheat grain are caused by attacks of the major insect pests, i.e., cereal leaf beetle (Oulema melanopus), sunny pest (Eurygaster integriceps), and aphids (Schizaphis graminum, Diuraphis noxia) (Pett and Muminjanov, 2006a).

Previous studies have identified increased effectiveness of seed production for cereal crops (Muminov, 2009) and improved wheat management technologies (Rashidov, 2009) as key measures for increasing the security of wheat production in Tajikistan. However, yield still remains low, calling for further concerted actions to increase yield simultaneously with securing environmental sustainability. This requires multi-faceted research to identify the major constraints within wheat production and sustainable ways forward to secure yield, and thereby food production in Tajikistan, without affecting the environment. No previous study has sought to provide insights into crop production in farmers' fields in Tajikistan.

The aim of the present study was thus to evaluate and define the major constraints affecting wheat yield and the sustainability (economic, social, and environmental) of wheat production in Tajikistan. Specific objectives were: (1) to gain knowledge about farming practices; (2) to assess the impact of biotic and abiotic factors on wheat yield; (3) to understand sustainability issues, including impacts on the economy, society, and the environment, of the Tajik wheat production system; and (4) to identify actions necessary to increase the yield and sustainability (economic, social, and environmental) of Tajik wheat production. The intention was to provide a multi-faceted understanding of the major constraints to increasing the yield of wheat in Tajikistan and of the concerted actions needed for achieving sustainability (high yield, and thereby good economic and social development, without hampering biodiversity and the environment) of wheat production, and securing high yield of good quality. Due to the importance of wheat as a staple crop world-wide, the findings of the study can also be applicable in actions to achieve a sustainable increase in production of staple crops in other developing countries. To enhance the opportunities for higher wheat yield, without increasing chemical inputs and decreasing biodiversity, and thereby hampering sustainability in the production system, farmers need better access to resistant, and high-yielding varieties. Therefore, in this study the major wheat varieties and advanced breeding lines in Tajikistan were screened for resistance to the most important diseases (common bunt, stripe rust, and leaf rust) and their potential as parents within the national breeding program was assessed.

A survey of farmers' fields under wheat production in Tajikistan was conducted in the period 2012–2014, in order to identify major constraints affecting production and sustainability. All the major small-grain cereals (i.e., wheat, barley, rye, and oats) were recorded in the survey, but the focus of the present analysis is on wheat, which is by far the most dominant crop in Tajikistan, occupying the significantly largest production area. In total, 210 fields were surveyed over the 3-years study period, with 53 wheat fields surveyed in 2012, 77 fields in 2013, and 80 fields in 2014. The surveyed fields included both irrigated and rainfed wheat fields, which were randomly chosen during each of the years, and therefore differed between years, to cover as many different farmers' fields as possible. The surveyed fields were divided according to district and agro-climate zone, as shown in Figure 1 and further defined in Table 1. Previous studies have also divided the country into agro-climatic zones, e.g., a system of 11 agro-climate zones devised by Babushkin (1964) is still widely used by local scientists (Saidmuradov and Stanyukovich, 1982). Later, simpler zone divisions have been suggested, e.g., the following four agro-climatic zones (Muminjanov et al., 2016): (i) Hot, dry, comparatively fertile valleys in the south; (ii) hot, dry, comparatively less fertile valleys in the north; (iii) cool and in some places humid zones in the foothills; and (iv) cold, dry, less fertile highlands in the east. The present survey was designed to cover the main areas where cereals are grown in Tajikistan, and therefore fields were classified based on their geographical and administrative location into seven zones (ZI-ZVII) (Figure 1 and Table 1).

Constraints and perspectives on sustainable wheat production in Tajikistan were identified in this survey. Farmers' knowledge, here measured as knowing which wheat variety they were growing (variable 3), was the outstanding factor influencing the performance of the crop, thereby affecting economic sustainability (Table 5). Lack of knowledge in this regard resulted in less good general performance of the crop, increased signs of drought effects, and increased levels of weed and diseases (variables 12, 14, 15, 17, 18, 20), all factors known to be correlated with decreased yield (Waddington et al., 2010; Amelework et al., 2016). Lack of knowledge was also related to low levels of crop rotation, indicating that farmers with little knowledge of the variety they were growing also lacked knowledge about the benefits of increased and secure yield through crop rotation. This lack of crop rotation might also have a negative environmental impact (Wienhold et al., 2006), thereby negatively affecting the environmental sustainability of wheat production. The low level of crop rotation and high degree of “wheat followed by wheat” cultivation in the Tajik system may be because wheat is a critical staple food for the Tajik people (Braun, 2003), but the wheat crop is relatively low-yielding in the country (FAO-REU, 2016). This low economic sustainability (low yield in the staple crop for food security) negatively affects the environmental sustainability (crop monoculture, as farmers need to grow wheat continuously in order to feed the population), and probably also the social sustainability (as communities get locked into trying to produce enough food to avoid hunger/starvation). Social sustainability is well-known to be connected to poverty and low educational status (UNESCO, 2019; UNICEF, 2019).

Lack of knowledge of the variety grown was most likely a result of farmers using farmer saved seeds, a behavior known to be common among farmers in Tajikistan (Husenov, 2018). Use of farmer saved seeds may be an indication of use of traditional landraces, but also of lack of income to buy seeds. Traditional landraces are known for their local adaptation and that some of them contribute specific characters e.g., they may be high in minerals, nutrient dense, tasty and often have a higher capacity to withstand weed pressure (Murphy et al., 2008; Hussain et al., 2010, 2012; Moreira-Ascarrunz et al., 2016). Appropriate use of farmer saved seeds with a careful cleaning of the seeds to get rid of e.g., weed seeds, can thereby contribute positively to sustainability, by local adaptation of the seeds, resulting in crop mixtures contributing to abiotic and biotic stress tolerance and also to securing agrobiodiversity in the crop, helping to counter future constraints (Constanzo and Bàrberi, 2014). However, farmer saved seeds can also spread seed-borne diseases and weed seeds from the previous harvest (Husenov et al., 2017). Previous studies have shown presence of spores of seed-borne diseases in practically all wheat material in Tajikistan (Husenov, 2018). In the present study, we found that in one-third of the fields surveyed, farmers are co-cultivating wheat with weed (variable 8), with weeds comprising more than 40% of the production (Table 2). One possible reason for that could be, e.g., negative impact of modern agricultural practices, like use of fungicides and herbicides, on landraces' performance. However, as only 5% of the farmers are using such chemical inputs, according to our study, this is not a likely explanation. Thus, the use of farmer saved seeds because of poverty, lack of resources to buy proper seeds, and lack of locally adapted landraces (Schmidt et al., 2019), are the most likely reasons for the high amount of weed on the fields, which can be a direct cause of inferior crop performance and yield, thereby contributing negatively to sustainability in crop production. However, without solving the weed issue, the use of modern wheat varieties might not be a better solution, except from such seed being free from weed, than the use of landraces, as modern wheat often is less weed competitive than lanraces (Murphy et al., 2008). Use of suitable wheat material to achieve high yield of desired quality can be seen as the second most important factor to increase yield, and thereby economic sustainability, in Tajik wheat production. The present survey revealed the negative impact of lack of knowledge of the variety grown on crop production (Tables 2, 3, 5) and lack of resistance in the Tajik wheat material (Table 6). The use of resistant and weed-competitive wheat material, obtained either from local landraces of traditional varieties (Rahmatov et al., 2019) or from novel varieties developed through modern breeding, could also contribute to continued low use of chemicals in the Tajik agricultural system, thereby contributing to environmental sustainability. However, the currently grown wheat varieties, and also the breeding lines screened in this study, showed limited levels of resistance, with most of the widely grown varieties showing no resistance to seed-borne diseases and all showing susceptibility to the three major diseases (Table 6), creating a higher risk of severe losses in the event of epidemics. Previous studies have found that home-grown wheat seed in Tajikistan contains high levels of seed-borne disease pathogens and has low protein quality, indicating both low level of resistance and low bread-making quality in the majority of the wheat produced in the country (Husenov, 2018).

Bread is the major staple food in Tajikistan, as confirmed by our finding that most wheat was produced for food purposes (variable 8) (Table 2). Wheat material from the national plant breeding program in Tajikistan has better quality properties than home-grown material (Husenov, 2018), although a comparison of current wheat varieties with novel advanced breeding lines found limited signs of improved baking quality (Husenov et al., 2015). Increased wheat yield is essential if bread continues to be the staple food in the Tajik diet. To achieve this, monoculture of wheat should be avoided, and the economic and social sustainability of wheat production should be raised. At present, wheat contributes 60% of the daily calorie intake of the Tajik population, but the country is not self-sufficient in wheat, resulting in Tajikistan ranking highly on the hunger map, while crop monoculture is known to be negative for agricultural sustainability (Braun, 2003; FAO, 2015; WFP, 2017). To increase wheat yield in Tajikistan, the total amount of weeds in wheat fields needs to be decreased. Relatively low amounts of weeds were found in one-third of the wheat fields surveyed here, but in another one-third of fields the weed density was >40%, meaning that the farmer grew almost as much weeds as wheat (variable 14) (Table 2). As mentioned above, farmer saved seeds might be one reason for this high weed density. The other reason might be the almost total absence of weed control (variable 10) (Table 2). In weed control strategies, the biodiversity of the flora always has to be taken into consideration. Tajikistan is known as one of the “cradles” of weed flora, as it is rich in endemic plants (Nowak et al., 2011, 2013). A high level of biodiversity is a sign of environmental sustainability of the Tajik agricultural system, despite monoculture of wheat. Moreover, the wheat varieties grown has been found to consist largely of variety mixtures of unknown origin (Husenov et al., 2015), thereby contributing to biodiversity in the field. To maintain biodiversity and environmental sustainability in wheat production, while simultaneously reducing weed density for economic and social sustainability reasons, education of farmers is crucial. Farmers need to be educated on coping strategies for maintaining high biodiversity, combined with weed control strategies to reduce weed pressure in production fields and thus sustainably increase yield in Tajik wheat production. Weed management techniques, such as mechanical control and crop rotation, combined with cultivation of allelopathic varieties able to compete with weeds for growing space, may contribute to higher yield, as may use of certified seed or clean seed. Seed breeding programs need to take into consideration novel varieties with specific resistance genes and combine them with locally adapted varieties with certain qualities and possibly also with variety mixtures, to sustain resistance and quality (Husenov, 2018).

A high percentage of the wheat fields surveyed contained >40% weeds (variable 14), of more than 18 different types (Table S1), so weed biodiversity is high in these wheat fields, which can be taken as a sign of sustainable wheat production. However, such a heavy weed density is known to reduce wheat yield drastically (e.g., Rasmussen, 2004), which is not socially or economically sustainable from a food security perspective (Smith and Gregory, 2013). Recent studies have shown that landsharing (combining fields for high-yielding crop production with fields for high biodiversity) can be the most beneficial solution, resulting in high crop production and good biodiversity (Colbach et al., 2018). Lack of crop rotation is another issue hampering sustainability in the Tajik cropping system (Wienhold et al., 2006), despite high biodiversity in the wheat varieties grown (Husenov et al., 2015). Due to lack of land and high demand for wheat from consumers, with wheat contributing 60% of daily calories (Braun, 2003), wheat is often grown in monoculture in Tajikistan, often also using farmer saved, home-grown seeds. Having wheat as the pre-crop for wheat (variable 2) was taken as an indicator of lack of crop rotation in the present study, but these systems were not found to be less sustainable (higher weed density, more diseases, etc.) than those with other pre-crops. Instead, a vague correlation between lack of farmer knowledge and lack of crop rotation was found (Table 4). However, the high reliance on wheat for food security might be an increasing risk factor under climate change with an increasing frequency of extreme weather events (Mukamuhirwa et al., 2019) and emergence of novel, highly devastating wheat diseases for which resistance genes are lacking (Rahmatov et al., 2016, 2019). Therefore, a more diversified crop cultivation would be beneficial for increased sustainability and food security in Tajikistan.

Several of the farmers interviewed reported drought problems in their fields and drought effects were observed in 18% of fields (variable 15), despite 73% of the farmers using irrigation (variable 11). Drought effects were observed in both irrigated and rainfed fields, and the reason reported by the farmers was lack of water access, especially in early spring.

The results showed variations in wheat crop performance between years, agricultural zones, altitudes, and farm sizes (Table 4). Previous studies have found larger farms to be more business-driven, with more commercial production and a more positive attitude to use of machinery, inputs such as pesticides, and high-tech solutions, e.g., genetically modified crops (Ahmed et al., 2011; Azadi et al., 2016). In the present study, larger farms were found to use more machinery (Tables 2, 4), but the farmers concerned did not have higher knowledge of e.g., what variety they were using and were not more successful in wheat production (Tables 2, 4), confirming earlier findings (Calviño and Sadras, 2002). Variations in crop production with agricultural zone and altitude are well-known from previous studies (Clay and Dejaegher, 1987; Hailu et al., 2010) and were confirmed in the present study. Figure 2 summarizes differences in measured variables by year of study, agricultural zone, altitude, and farm size, applying a PCA. As seen, the studied farms in ZI-ZII seemed to share some characteristics; they were generally small and situated <700 masl (Figure 2A), but similarities in cultivation issues were more difficult to find. Instead, altitude of the farm seemed more relevant, with common features for farms situated <700 masl being more common use of technical crops as preceding crops, low knowledge among farmers, production of wheat for food purposes, visible effects of drought despite use of irrigation, and low levels of loose smut and stripe rust in wheat fields (Figure 2B). Zones ZV and ZVII were found to have the most deviating features. In ZV, farms were in general of medium size and situated at 700-1000 masl (Figure 2A), and were characterized by the highest amount of knowledge among farmers, the lowest weed density in fields, and the greatest area of winter wheat, sown using machinery and for seed purposes (Figure 2B). In ZVII, the farms were situated at >2,000 masl (Figure 2A) and farmers used hand broadcasting for sowing spring wheat at the optimal time (Figure 2B). The wheat was grown for food, irrigation was used, and lodging was a problem (Figure 2B). Variations between years found in this study could be partly explained by the farmers' fields being selected independently for survey in each year, i.e., they were not the same over the years. However, other studies (e.g., Calviño and Sadras, 2002; Verón et al., 2004) have shown an impact of various constraints on yearly variations in yield. The variation over years for the different variables evaluated in this study might be the result of weather variations or outbreak of diseases etc., apart from the effect of farm selection.

As shown in previous studies (Sadras et al., 2002; Pretty et al., 2003), we found that improved farming practices, including knowledge of the farmer, use of a suitable variety, use of fertilizer, weed control, and irrigation, contributed to a healthier crop (Table 2), which is known to correlate with increased yield. Resistance genes to the major diseases were in principle lacking in the Tajik wheat material screened here (Table 6). Among the five most widely grown wheat varieties in 2011 (Krasnodar 99, Basribey, Starshina, Lastochka, and Steklovidnaya 24) (see Table 3), together representing about 57% of the area under wheat (Muminjanov et al., 2016), two varieties (Krasnodar 99 and Steklovidnaya 24) showed high susceptibility to all three diseases tested, while the others were susceptible to one or more of these diseases. Breeding tolerant varieties that are also competitive to weeds has not even started. Strategies to secure biodiversity and simultaneously increase yield with continuous low chemical input are in place to a limited extent. Thus, our results clearly show that, to increase sustainable wheat production in developing countries, where Tajikistan can be seen as an example, concerted action is urgently needed, with education of farmers and suitable incentives for steering the development of farming systems in a sustainable direction being of critical importance.

Three key major constraints to sustainable high-yielding wheat production in Tajikistan were identified: (i) lack of knowledge among farmers (lack of knowledge about variety grown, correlating with lack of crop rotation and poor crop performance); (ii) lack of use of certified seed and resistant varieties in production (lack of crop knowledge, related to use of saved seeds and lack of resistance in wheat varieties grown); and (iii) lack of suitable management systems, in particular for weeds (co-production of wheat and weeds, and no weed management). Differences in wheat production also arose from variations in cropping practices between regions (zones), altitudes, and farms of different sizes, and from selection of different farms for survey in different years, but these variations were lower than those caused by the three key factors.

A combination of greater knowledge among farmers, use of certified seed and resistant varieties, and suitable weed management could contribute to higher yield and increased social and economic sustainability (higher yield of the main staple, contributing to less hunger). Increased wheat yield would improve the chances of Tajikistan becoming self-sufficient in wheat, the staple food of the country, reduce hunger among local people, and provide opportunities to grow other crops to support the household with food, which might lead to a change from wheat monoculture to a more diverse system. Higher yield might also create economic opportunities for actions to secure high biodiversity in e.g., weed species in the country. Sustainable weed management techniques, such as crop rotation, tolerant wheat varieties, and e.g., landsharing, need to be developed for the country to secure the current high biodiversity and limit chemical solutions.

Thus, to secure sustainability (environmental, economic, and social), the Agriculture Authority of Tajikistan needs to educate farmers in wheat production, make certified seed widely available, and develop suitable weed management systems, while also maintaining high biodiversity.

All datasets generated for this study are included in the article/Supplementary Material.

BH, EJ, and HM planned the study, the hypothesis, and the objectives. BH and SA developed the questionnaire. BH and MO conducted the survey, with guidelines by AM and SA. AM assisted in screening of varieties in Turkey. BH, EJ, and LG-G evaluated the results. EJ and HM provided overall coordination for the project. BH drafted the manuscript, together with LG-G. All authors contributed to manuscript improvement, read, and approved the final manuscript.