The study was carried out in the Kanchanpur district of Nepal, situated at 28°12'N, 82°10'E, characterized by a climatic environment conducive to wheat production. Within the Kanchanpur district, the Bhimdatta municipality, Bedkot municipality, and Dodhara Chandani municipality were deliberately selected, given their substantial involvement in wheat farming and their classification within the PMAMP wheat zone Kanchanpur. The study involved the random selection of one hundred wheat farmers through a simple random sampling method from the total wheat growers within the PMAMP region. Face-to-face interviews were conducted with the selected farmers’ households utilizing meticulously prepared and pre-tested semi-structured questionnaires. The research was performed during the months of March and April in the year 2022.

According to an analysis of the minimum, maximum, and average standard deviation of land holdings within sampled farmers’ households, a classification into three distinct groups was undertaken. Specifically, households with land holdings ranging from the minimum to the mean standard deviation were classified as small household farmers. Similarly, households with land holdings ranging from the mean minus standard deviation to the mean plus standard deviation were categorized as medium farmer households. Lastly, households with land holdings ranging from the mean plus standard deviation to the maximum range were classified as large farmer households. The findings indicated that among the 100 farmers included in the study, 37 were identified as belonging to the small farmers group, 41 fell within the medium farmers group, and 22 were categorized as large farmers. The average land holdings of these respective groups were calculated as 10.20 kattha for small farmers, 17.39 kattha for medium farmers, and 41.9 kattha for large farmers.

The data acquired through face-to-face interviews was tabulated using MS-Excel v2019 and subsequently imported into Python 3.8 for subsequent analysis. Descriptive analysis was undertaken, specifically categorizing farmers into small holders, medium holders, and large holders. The comparison of continuous variables was executed through one-way analysis of variance (ANOVA), while the comparison of categorical variables was conducted using a contingency chi-square test, all with respect to different farmer categories.

The factors influencing wheat production were examined through the utilization of a multiple regression model, following the framework proposed by Thompson (1978).

Ln (Yproduction) = α+β1 lnX1+ β2 lnX2+ β3 lnX3+ β4 lnX4+ β5 lnX5+ β6 lnX6+ β7 X7+ β8X8

α = intercept made on the regression line

X1 = Number of economically active members in the respondent household.

X2 = Total wheat cultivated land with irrigation facility

X3 = Experience of the household head in wheat cultivation (in years of wheat cultivation)

X4 = Education level of the respondent household head

X5 = Total Chemical Fertilizer (NPK) applied

X6 = Seed Rate

X7 = Dummy for intercropping if yes=1 and if no=0

X8 = Dummy for training in wheat cultivation if yes=1 and if no=0

β1 to β6 are the coefficients of those independent variables respectively.

Six continuous variables and two dummy variables were chosen as the independent variables in this model. The intercorrelated independent variable was eliminated, and the variable having the greatest impact on production was chosen, preventing multicollinearity in the model. This methodology aligns with the approach used in analyzing factors and major problems in paddy farming production in Sunsari district, Nepal (Acharya et al. 2020). A similar model was applied to ascertain factors influencing productivity and major constraints in mango production in the Saptari district of Nepal (Shrestha et al. 2021). Furthermore, this method was employed to identify the productivity and profitability of vegetable production in Swaziland (Masuku and Xaba 2013).

The assessment of wheat production challenges was conducted utilizing a forced ranking technique employing a five-point scale. The determination of the index for the intensity of production problems faced by wheat growers was achieved through the application of the following formula:

I_imp = $\sum =\frac{Sifi}{N}$

Where,

I_imp = index of importance

∑ = summation

S_i = I^th scale value

F_i = frequency of i^th importance given by the respondents

N = total number of respondents

A similar formula was employed to rank the challenges associated with wheat production (Subedi et al. 2019). This methodology aligns with the approach adopted by Acharya et al. (2020) in ranking the production constraints of rice in the Sunsari districts of Nepal, who similarly utilized a five-point scale. Furthermore, Shrestha et al. (2021) employed a similar five-point scale to rank the production constraints in mango farming in the Saptari district of Nepal.

The sampled respondents’ socioeconomic characteristics are displayed in Table 1. The average age of the head of a household was 41.22 years for small farmers, 47.51 years for medium farmers, and 47.09 years for large farmers. For small farmers, the average household size was 5.65; for medium farmers, it was 6.32; and for large farmers, it was 7.55. For small, medium, and large farms, the average number of males in the household was 2.72, 3.26, and 4.13, respectively. Similarly, for small farmers, the average number of females in the household was 2.91; for medium farmers, it was 3.04; and for large farmers, it was 3.40. The age composition of the surveyed household population was stratified into distinct categories: 0–5, 6–17, 18–59, and 60 and above. Within this framework, individuals falling within the age range of 18–59 are regarded as the economically active population, while those in other age groups are classified as the dependent population.

The average count of the dependent population was determined to be 2.37 for small farmers, 2.45 for medium farmers, and 3.10 for large farmers per household. Likewise, the average count of the economically active population was observed to be 3.64 for small farmers, 4.10 for medium farmers, and 4.86 for large farmers per household. The dependency ratio, defined as the proportion of the dependent population to the economically active population (Ramshah Path 2014), was computed as 0.82 for small farmers, 0.69 for medium farmers, and 0.74 for large farmers. A dependency ratio of 0.82, 0.69, and 0.74 implies that, on average, 100 small, medium, and large farmers must sustain 82, 69, and 74 individuals, respectively, who are dependent on them.

Furthermore, the average landholding size was found as 10.20 kattha, 17.39 kattha, and 43.90 kattha for small, medium, and large farmers, respectively.

The socio-demographic data encompassing the gender of the household head, educational attainment, religious affiliation, family structure, and primary occupations are detailed in Table 2. Predominantly, the respondents were male, except for the small farmers’ group where 32.4% of respondents were male. This difference was statistically significant at a 1% level of significance, indicating a noteworthy contingency in gender distribution across farmer categories. The major religion identified across all farmer groups was Hinduism, and this data did not yield statistical significance at any level.

Ethnicity exhibited notable variations, with Brahmin and Chettri being predominant among small and medium farmers, while Janjati held dominance in the large farmer group. The ethnic disparities were statistically significant at a 5% level. Wheat cultivation emerged as a predominant occupation in all farmer groups, demonstrating statistical significance at a 5% level. The educational profile indicated a prevalence of illiteracy among small farmers, a substantial proportion of middle farmers having passed the School Leaving Certificate (SLC), and a predominant SLC level of education among large farmers. However, these differences were not found to be statistically significant.

Regarding training on wheat cultivation, 27% of small farmers’ households, 29.3% of medium farmers’ households, and 59.1% of large farmers’ households had received such training at least once. This discrepancy was found to be statistically significant at a 5% level, partially attributed to the larger cultivation lands of large farmers, necessitating a greater emphasis on training due to their dependency on agriculture for income.

Regarding household decision-makers, the majority of decisions were made by females in small and medium farmers’ households (70.3% and 53.7%, respectively), while in large farmers’ households, males predominantly took major household decisions (68.2%). This difference was statistically significant at a 5% level.

Based on the data presented in Table 3, the R-square value was determined to be 0.95, indicating that approximately 95% of the variability in wheat production was elucidated by the explanatory variables incorporated in the model. Furthermore, the adjusted R-square value was calculated as 0.90, signifying that when accounting for degrees of freedom, approximately 90% of the variability in the dependent variable (production) was explicated by the independent explanatory variables within the model.

The table underscores that wheat-cultivated land equipped with irrigation facilities, the quantity of NPK applied, the years of experience, and the education level of the farmers made positive contributions to wheat production in the study area. Conversely, negative coefficients were observed in relation to economically active members of the farmers, seed rate, and intercropping. This result indicates that not all members of the working class actively participate in wheat farming within the household, primarily due to their engagement in other works such as government employment, business etc. and the results of Shehu et al. (2010) contradicted this finding, finding that an increase in the number of adult family members also increased yam production in Benue state, Nigeria. Similarly, the negative coefficient of seed rate suggests that as the seed rate increases, the total production of wheat tends to decrease. It indicates that using higher seed rates might lead to overcrowding of plants, which can result in competition for resources like water, nutrients, and sunlight. As a result, individual plants might not have enough space and resources to grow optimally, leading to reduced overall production, and a similar result was found by Khan et al. (2000). Their finding revealed that, due to competition among plants for nutrients, light, and air, the optimal biological yield was not achieved with the highest seed rate. Behzad and Amani (2020), conducted additional research which revealed that a seeding rate of 140 kg ha^-1, as opposed to 160 kg ha^-1, produced the maximum grain production (4.94 t ha^-1).

The negative coefficient associated with the intercropping of green pea with wheat signified that the standard ratio of wheat and green pea were not followed and a huge number of green pea were intercropped which has a detrimental effect on yield. A high ratio of wheat and green pea leads to overcrowding of plants, which results in competition for resources like water, nutrients, and sunlight rather than being complementary and this result was in contrast with Dhillon and Aulakh (2017) as they found that wheat intercropping with chickpeas result on higher production of wheat (54.6 q ha) than sole cropping of wheat (46.7 q ha). Similarly, Hamdollah and Ghanbari (2010) showed that intercropping results in high productivity and profitability.

The presence of wheat-cultivated land equipped with irrigation facilities exhibited a positive and statistically significant impact on wheat production, evident at a 1% level of significance. This observation indicated that, for each additional unit increase in wheat-cultivated land with irrigation, there was a corresponding 0.617 unit increase in wheat production. Specifically, this implies that incresing the wheat-cultivated land by an additional 1 kattha, with an irrigation facility, resulted in an increase of wheat production by 0.617 quintal, holding other factors constant. A similar outcome was noted by Acharya et al. (2020), where an increase of 1 kattha in paddy-cultivated land with irrigation facilities led to a 0.0108-ton increase in paddy production, under constant conditions. Moreover, Zaveri and Lobell (2019) reported that India’s national wheat yields in the 2000s were 13% greater than they would have been in the absence of irrigation patterns since 1970. According to Steiner et al. (1985), irrigation had a major impact on the wheat crop’s dry matter production, grain yield, and yield components. It also took up more than 70% of the crop’s water requirements during arthesis. It has been shown that more frequent irrigation produces higher grain yields than less frequent irrigation in several nations (Singh and Brar 1979).

The amount of chemical fertilizer NPK applied had a positive impact on the total production of the Wheat. It was found statistically significant at a 1% level of significance. With a one-unit increase in the fertilizer, it increases the production by 0.054 quintal other things remain constant. Fertilizers need to be applied to wheat at three different times i.e. at tillering, booting, and at grain filling. Maximum growth metrics responded considerably to NPK fertilizers, according to Malghani et al. (2010). It is found that the application of 175-150-125 NPK Kg ha^-1 resulted in the highest grain yield, which was 5168 Kg ha^-1. When comparing the increased yield to the control (2502 kg ha^-1), which did not get any fertilizer, the difference was 51.58%. The greatest production stability and lowest agronomic risk for yield failure were offered by the mineral supply of N + P + K plus extra manure (Macholdt et al. 2019).

The acquisition of training by farmers in wheat production demonstrated a favorable effect on wheat production; however, its influence did not attain statistical significance concerning the overall wheat production within the study area. This lack of significance can be attributed primarily to the diminished efficacy of the training program implemented in the study area. Notably, this result diverges somewhat from the findings of Kijima et al. (2012), where training was identified to have a statistically significant positive impact on the adoption of improved practices and the profitability of lowland wheat production in Uganda. Ahmad et al. (2007) found that due to training, crop yield has increased, vegetable and fruit production has also shown an upward trend, and the diseases and mortality rate of the livestock has also decreased. Wonde et al. (2022) found that trainees increased their wheat and maize yield by 26.66% and 10.10% kg ha^-1, respectively with a net annual income increase of 19.64% from wheat production in Ethiopia.

The educational attainment of the respondent household head exerted a positive influence on wheat production. This observation suggests that individuals with higher levels of education possess greater awareness of cropping systems and production practices, enabling them to more readily adopt innovative agricultural technologies. Ullah et al. (2014) concluded that education has a positive impact on the productivity of wheat. The farming skills and productive capabilities of the farmers were enhanced by education (Weir 1999). Paltasingh and Goyari (2018) observed a positive correlation between the education level of farmers and farm productivity among Indian farmers. However, Arshad et al. (2015) reported a contrasting negative impact of education on flood insurance in rural households in Pakistan.

In the context of wheat cultivation, the experience of farmers quantified in terms of years, demonstrated statistical significance at a 10% level. The coefficient value of 0.166 indicates that with other variables held constant, a one-unit increase in experience resulted in a corresponding 0.166-unit increase in wheat cultivation in the study area. This aligns with the findings of Indian farmers (Coelli and Battese 1996). Gedara et al. (2012) also observed a positive and significant impact of farmer experience on efficacy. This concurrence is further supported by the work of Wilson et al. (2001) where the managerial experience of farmers was linked to an increased quest for new information and, subsequently, higher technical efficiency in wheat production in Eastern England.

Some of the main issues that are common in the study area were shortlisted and put in the interview schedule based on the field tour, focus group discussion, knowledge sharing between KII and AKC, and wheat zone. Following these discussions, key issues such as the unavailability of fertilizers during critical wheat plantation periods, damage caused by wild animals and stray cattle, lack of irrigation facilities, the prevalence of diseases and insect pests, insufficient access to markets, absence of quality seedlings, unavailability of training for commercial wheat production, and the lack of governmental support for agricultural inputs emerged as significant impediments to successful wheat farming. To rank these challenges, a forced ranking methodology utilizing a five-point scale was employed, prompting farmers to assign a maximum score of 1 and subsequently decreasing the score in proportion to the perceived severity of the problem. Notably, the majority of farmers identified the lack of fertilizers during critical plantation periods as a severe issue, attributing it to the limitation of the soil’s overall production capacity and hindrance to plants reaching their maximum potential, resulting in diminished yields due to the essential nutrient deficiency crucial for growth, development, grain formation, and grain filling. The calculated index value for the fertilizer shortage was determined to be 0.795. Similarly, issues such as damage from wild animals and stray cattle, unavailability of improved seed, unavailability of irrigation, prevalence of diseases and insect pests, insufficient access to markets, labor shortage, lack of training on improved farming practices and natural hazards were ranked 2^nd, 3^rd, 4^th, 5^th, 6^th, 7^th, 8^th, and 9^th, respectively, with index values of 0.790, 0.718, 0.695, 0.647, 0.470, 0.336, 0.316, and 0.225. The study area is a conservation area where Sukhlapantha National Park lies, the damage from wild animals is a serious problem. Not only wildlife but also damage from stray cattle are also serious problems for farmers.

The third problem lack of quality seeds involves the unavailability of high-yielding varieties of wheat. Consequently, farmers face challenges in commercializing their production. Farmers rely on the local variety or stored seeds of previous years which are highly vulnerable to disease and pests which result in low productivity of wheat. As a result, farmers’ income has decreased and production costs have increased.

Despite the freely flowing Mahakali River near the study area, there was a problem of irrigation. The huge potential of the Mahakali river hadn’t been exploited yet. The only available source of irrigation was through water pumps and small canals. Only very few farmers had all-year-round access to irrigation. Moreover, being on the leeward Western side of Nepal this region received very minimal precipitation. Some of the farmers responded that there was a shortage of drinking water as well, irrigation meanwhile for the cultivation was secondary. Infestation of the diseases and pests in the field caused a severe loss in production. Leaf rust, Fusarium head blight, Septoria leaf blotch, stripe rust, spot blotch, tan spot, and powdery mildew are the fungal pathogens that cause the most severe losses. Insects such as White grubs, wireworm, fall-army worm, cutworm, and wheat stem sawfly causes major losses in wheat.

Insufficient access to the market was ranked the sixth important problem which involves the improper amount received by the farmers. As a result of the syndicate upon the processing of wheat, the farmers are not able to get the proper price of wheat. The large-scale farmers with huge capital have a monopoly over the processing plant which causes lower wheat for the produce of small farmers. No access to training was also ranked as a problem. The migration of people in search of jobs in India causes the scarcity of labor. Natural hazards like drought, hailstone, and flooding were ranked least important by the farmers.