Socio-economic characteristics, landholding patterns and tree species composition in agroforestry systems of the indo-gangetic plains: A case study from Fatehpur, Uttar Pradesh, India

Introduction

Agroforestry -the deliberate integration of trees with crops and, in many systems, livestock -is widely regarded as a holistic approach to agricultural sustainability, combining resource-use efficiency, livelihood diversification and environmental benefit within the same land unit (Raj et al., 2019) [11] . On-farm trees supplement staple-crop income, provide fodder and fuel, and contribute to livelihood resilience among smallholder farmers through diversified farm outputs and resource use (Debta, 2025; Palsaniya et al., 2023) [3,9] , with homestead and farm-boundary tree systems shown to secure a more consistent household product supply and generate familylabour employment (Singh et al., 2016) [14] . The Indo-Gangetic Plains (IGP) -spanning Punjab, Haryana, Uttar Pradesh, Bihar and West Bengal -are a priority region for agroforestry integration, combining a dense agrarian population with extensive cereal-based cropping systems into which tree components can be incorporated (Ram et al., 2022) [12] . Geospatial suitability mapping identifies the IGP as a high-priority zone for tree-based systems nationally (Ahmad et al., 2019) [1] , and enterprise-mix diversification incorporating trees alongside staple crops has been proposed to improve the ecological sustainability of smallholder livelihoods in the region (Yadav et al., 2022) [17] . Evidence from eastern rural India further indicates that sustainable agricultural practices can be associated with improved multidimensional livelihood outcomes (Khosla et al., 2025) [6] . Uttar Pradesh illustrates the diversity of agroforestry pathways within the IGP. Western UP has undergone a marked shift toward commercial, poplar-and eucalyptus-based agroforestry since the 1990s, while central and eastern UP have retained more traditional, multi-species tree management (Verma et al., 2017) [16] . Farmer surveys from Ballia district, eastern UP, document a predominance of scattered farm and homestead trees, with fruit, timber, shade and fodder ranked among the principal benefits (Shukla et al., 2020) [13] , and species inventories from the Vindhyan region, bordering Fatehpur, report a timber-dominated, multi-species assemblage alongside a demand-supply gap for quality timber (Srivastav et al., 2024) [15] . Education showed the most consistent association with landholding. In a survey of 176 western-UP households, landholding size, irrigated-land area and prior tree-planting experience were positively and significantly associated with the decision to plant trees, whereas household size and education were not (Pande, 2021) [10] . Because farm-forestry in India is practised predominantly by small and marginal landholders raising trees on bunds and field boundaries, livelihood outcomes depend on integrating trees within, rather than displacing, existing cropping systems (Pande, 2021; Kashyap et al., 2022) [10,5] . Many species farmers retain are not single-purpose: UP agroforestry inventories repeatedly show a share of species serving more than one household function simultaneously, combining, for example, timber value with fruit or fodder yield (Srivastav et al., 2024) [15] . Comparable functional diversity has been documented in an integrated agroforestry characterization from Telangana (Naveen et al., 2025) [7] -a markedly different agro-climatic setting, but one illustrating the general pattern of multipurpose species retention by smallholders. Documenting which functions are served by which species is therefore central to understanding the practical utility of farmer-managed tree inventories, beyond a simple species list. Several studies have characterised parts of this picture elsewhere in UP, but each has addressed only a component of it for the central-UP landscape in which Fatehpur district sits. Socio-economic and land-use patterns have been documented for western UP (Verma et al., 2017; Pande, 2021) [16,10] , Ballia district (Shukla et al., 2020) [13] and Sonbhadra district (Anand et al., 2016) [2] ; a gender-focused survey in adjacent Sultanpur district has separately documented socio-economic decision-making patterns and a partial species list (Yashmita-Ulman & Patel, 2025) [18] ; and tree-species composition has been documented in more detail for the Vindhyan region (Srivastav et al., 2024) [15] . An integrated treatment combining socio-economic and species-functional data has been reported for Telangana (Naveen et al., 2025) [7] , and standardized socio-economic instruments have recently been proposed for South Gujarat (Deshmukh et al., 2023) [4] illustrating that integrated, statistically grounded characterization is achievable, though not yet reported for a central Uttar Pradesh, IGP setting. The existing literature has therefore documented individual components of agroforestry systems in Uttar Pradesh -socioeconomic characteristics, landholding patterns, or tree-species composition -but their integration with land-source/tenure and farmer-reported functional classification of tree species, within a single respondent-level dataset, remains limited for the central Uttar Pradesh context. The present study addresses this gap by combining, within a single dataset from two villages of Fatehpur district, (i) household socio-economic characteristics, (ii) landholding size and land-source (tenure) patterns, (iii) statistical associations between socio-economic characteristics and landholding, and (iv) a tree-species inventory classified by farmer-reported functional category (Timber, Fruit, Fodder), with explicit identification of species recorded in more than one category. The specific objectives of this study were to characterize: (i) the socio-economic characteristics of surveyed agroforestry households in Sultangarh and Sijoili villages, Fatehpur district; (ii) landholding size and land-source (tenure) patterns among these households; (iii) statistical associations between household socio-economic characteristics and landholding; and (iv) the composition of the tree-species inventory maintained within these agroforestry systems, including the functional overlap of species recorded in more than one farmer-defined category (Timber, Fruit, Fodder).

Materials and Methods

Study area

The study was conducted in Sultangarh and Sijoili villages of Bindki tehsil, Fatehpur district, Uttar Pradesh, India, within the central Indo-Gangetic Plains. Fatehpur district, one of the four districts of the erstwhile Prayagraj division, occupies approximately 4,152 km² and lies in the eastern (lower) part of the Ganga-Yamuna doab, between latitude 25°26′ N and longitude 80°14′-81°20′ E. Among the district's three tehsils -Fatehpur, Khaga and Bindki -Bindki was selected for this investigation. The district is drained by the Ganga river system (tributaries: Yamuna, Rind); soils are predominantly Gangetic alluvium (sandy loam to clayey), and the climate is sub-humid, with mean annual rainfall of 932 mm (~90% from the south-west monsoon) and mean monthly temperatures ranging from ~8°C in January to 46.5°C in May.

Study population, sampling, variables and tree-species classification

Data were collected from farm households practising agroforestry in Sultangarh and Sijoili, two villages of Bindki tehsil. A list of farmers was compiled for each village, from which 50 respondents per village were selected using a simple random sampling procedure, giving a total sample of n = 100 farmers, surveyed during the 2020-21 production season through a structured questionnaire administered by personal interview. For each respondent, the questionnaire recorded gender, age, educational attainment, source(s) of land (Family, Purchased, Leased, or combinations thereof) and total landholding area (acre; 1 acre = 0.405 ha). Alongside the household survey, tree species maintained within the agroforestry systems of the two villages were recorded and consolidated into a single inventory of common and botanical names, grouped into three farmer-reported functional categories -Timber, Fruit and Fodder. Species reported under more than one category (e.g., Neem, Siris, Mahua, Subabul and Babul under Timber and Fodder; Jackfruit and Jamun under Timber and Fruit) were retained in every category reported, yielding 16 timber, 7 fruit and 5 fodder entries -28 functionalcategory entries corresponding to 21 unique tree taxa overall.

Data processing and statistical analysis

Raw data were cleaned and categorical entries standardized (gender, education and land-source labels). Four variables were derived: age class (Young <35, Middle 35-55, Old >55 years); education score (ordinal, Illiterate = 0 to Graduate = 4); farm category, based on landholding (Marginal <2.5 acre, Small 2.5-5.0, Medium 5.0-10.0, Large >10.0 acre); and land source (Owned:

Family/Purchased/Family+Purchased; Mixed: Family+Leased; Leased: Leased only). Data cleaning, derivedvariable computation and statistical analyses were performed in Python (pandas, SciPy). Categorical variables were summarized using frequencies and percentages, while continuous variables were described using mean, median, SD, range and coefficient of variation. Associations among categorical variables were assessed using the chi-square test, monotonic relationships using Spearman rank correlation, and differences in landholding distributions among groups using the Kruskal-Wallis H test. Statistical significance was set at α = 0.05.

Results and Discussion

Socio-economic characteristics of surveyed households

Of the 100 respondents, 86.0% were male and 14.0% female (Table 1; Figure 1a), consistent with male-headed respondents predominating in comparable Ballia-district surveys (Shukla et al., 2020) [13] . With only 14 female respondents, the nonsignificant gender-farm-category association (χ² = 0.735, df = 3, p = 0.865; Table 3, Panel A) should not be read as evidence of gender parity in land access; a gender-focused survey in adjacent Sultanpur district similarly documents pronounced male dominance in agroforestry decision-making (Yashmita-Ulman & Patel, 2025) [18] . The age-class distribution was skewed toward older respondents -Old (>55 years) 52.0%, Middle (35-55) 39.0%, Young (<35) only 9.0% (Table 1; Figure 1b). The predominance of older respondents is broadly comparable with observations from agroforestry households in Ballia district (Shukla et al., 2020) [13] . Educational attainment was concentrated in the lower categories: Illiterate (27.0%) and Primary (32.0%) together accounted for 59.0% of respondents, against 21.0% High School, 15.0% Intermediate and 5.0% Graduate (Table 1; Figure 1c). This is broadly consistent with lower rural literacy levels reported for Uttar Pradesh (Verma et al., 2017) [16] ; socio-economic factors including education and age have similarly been shown to shape agroforestry adoption in neighbouring UP districts (Anand et al., 2016) [2] . Agroforestry was represented across all educational categories, indicating that the practice was not restricted to a particular level of formal education, as also reported for Mirzapur district, UP (Nayal et al., 2024) [8] .

Landholding patterns and land sources

Farm-size distribution was dominated by smaller holdings: Marginal (<2.5 acre, 39.0%) and Small (2.5-5.0 acre, 41.0%) together accounted for 80.0% of households, with Medium (5.0-10.0 acre, 17.0%) and Large (>10.0 acre, 3.0%) less common (Table 1; Figure 2a), mirroring the national predominance of small and marginal holdings and the bund/boundary-planted nature of Indian farm-forestry (Pande, 2021; Kashyap et al., 2022) [10,5] . Purchased land was rare, occurring alone in 2.0% of households or in combination with family land in 10.0% (Table 1; Figure 2b).

Combining categories, 59.0% of households held some leased component. Tenure security is repeatedly identified as a determinant of tree-planting behaviour in the Indian agroforestry literature (Verma et al., 2017; Pande, 2021) [16,10] . The prevalence of mixed and leased holdings highlights the importance of considering land-tenure structure when evaluating smallholder agroforestry systems. Mean landholding increased from 1.58 acre in marginal farms to 16.53 acre in large farms, while the pooled mean was 3.68 acre (median 3.00, SD 3.01, CV 81.8%, range 0.5-19.6 acre) (Table 2) 1. CV = coefficient of variation (SD/mean × 100). Overall row is the pooled sample statistic (n = 100), not a mean of the four category means. 1 acre = 0.405 ha.

Relationships between socio-economic characteristics and landholding

Farm category showed an apparent association with education (χ² = 27.599, df = 12, p = 0.0063) and age class (χ² = 16.377, df = 6, p = 0.0119) by chi-square, though both results are indicative rather than confirmatory given that >50% of contingency-table cells had expected counts <5 (55% and 50%, respectively; Table 3, Panel A). Farm category was not significantly associated with gender (χ² = 0.735, df = 3, p = 0.865), nor was education with land source (χ² = 22.391, df = 16, p = 0.131; 64% of cells <5). Landholding distributions differed significantly across education groups (Kruskal-Wallis H = 18.81, df = 4, p = 0.0009) and landsource groups (H = 16.919, df = 4, p = 0.0020) but not across ageclass groups (H = 5.829, df = 2, p = 0.0542) (Table 3, Panel B). Landholding was weakly and negatively correlated with age (Spearman ρ = -0.248, p = 0.0128) and weakly and positively correlated with education score (ρ = 0.241, p = 0.0157); age and education score were themselves very strongly and negatively correlated (ρ = -0.731, p < 0.0001; Table 3, Panel C). Education showed the most consistent association with landholding across all three approaches -a pattern reported elsewhere in Indian agroforestry contexts and attributed to mechanisms such as intergenerational transmission of education and land assets (Nayal et al., 2024; Shukla et al., 2020; Verma et al., 2017) [8,13,16] . Age showed a weaker, more equivocal relationship: because chi-square does not indicate direction, and the only directional test (Spearman) shows younger respondents holding slightly more land, the data do not support an interpretation of age-related land accumulation; the negative association therefore should not be interpreted as evidence of an age-related land-accumulation process. Because age and education are so strongly interrelated, their associations with landholding cannot be treated as independent effects from these bivariate tests.

Tree species composition and functional classification

The consolidated inventory (Table 4) recorded 21 unique tree taxa represented by 28 farmer-reported functional-category entries, comprising 16 Timber (57.1%), 7 Fruit (25.0%) and 5 Fodder (17.9%) entries (Table 5; Figure 3). The difference between the number of unique taxa and functional-category entries resulted from the classification of several taxa under more than one functional category. The timber-dominated composition was comparable with that reported from the neighbouring Vindhyan region of Uttar Pradesh (Srivastav et al., 2024) [15] , while differing from the more commercially specialised poplar-and eucalyptusoriented agroforestry systems reported from western Uttar Pradesh (Verma et al., 2017) [16] . Madhuca longifolia Note: † Recorded in more than one functional category: all 5 fodder species are also timber species, and 2 timber species (Jackfruit, Jamun) are also fruit species -7 of 28 entries cross-listed, giving 21 unique tree taxa overall (Table 6). Ber, Eucalyptus and Bamboo are recorded to genus level only, reflecting the resolution of the original field inventory.

The presence of a diverse fruit component, including mango, guava, aonla, bel and ber, also indicates the incorporation of treebased products beyond timber, consistent with the multifunctional nature of agroforestry systems documented in Uttar Pradesh (Shukla et al., 2020) [13] . A notable feature of the inventory was the multifunctionality of several tree taxa. Seven of the 21 unique taxa (33.3%) were reported under more than one functional category (Table 6). The largest overlap occurred between Timber and Fodder, with all five fodder taxa-Subabul, Babul, Neem, Siris and Mahua-also classified as Timber. This pattern indicates the importance of multipurpose tree classification in smallholder agroforestry, where individual taxa may be associated with more than one farmer-reported function (Srivastav et al., 2024) [15] . A smaller Timber-Fruit overlap was observed for Jackfruit and Jamun. No taxon was reported simultaneously under all three categories, and no direct overlap occurred between Fruit and Fodder. Thus, functional overlap was concentrated primarily at the Timber-Fodder interface, followed by the Timber-Fruit interface, indicating that farmer-reported tree composition was characterised by a combination of timber-oriented and multipurpose tree resources rather than by strictly single-purpose species groups.

Conclusion

This study characterized agroforestry-practicing households in two villages of Fatehpur district, Uttar Pradesh, integrating socioeconomic profiles, landholding and tenure patterns, and treespecies functional composition within a single dataset. Surveyed households were predominantly marginal and small farmers (80.0% of the sample) operating under tenure that was frequently mixed or leased (59.0% with some leased component). Educational attainment was the socio-economic characteristic most consistently associated with landholding across chi-square, Kruskal-Wallis and Spearman analyses, whereas age showed only a weak, negative association with landholding that should be interpreted cautiously given the strong collinearity between age and education in this sample. The tree-species inventory comprised 21 unique tree taxa dominated by Timber (16 of 28 functional-category entries, 57.1%), with Fruit (25.0%) and Fodder (17.9%) also represented; seven species (33.3%) served more than one functional role, chiefly at the timber-fodder interface. These findings describe a land-constrained, mixed-tenure smallholder agroforestry system with a functionally diverse, timber-dominant tree inventory. Agroforestry extension programmes in this landscape should account for the predominance of small and mixed-tenure holdings and the relatively low educational attainment of many practitioners. Extending this integrated characterization across additional central Uttar Pradesh districts would help establish how representative this pattern is of the broader Indo-Gangetic Plains.

References

1. Ahmad F, Uddin MM, Goparaju L. Agroforestry suitability mapping of India: geospatial approach based on FAO guidelines. Agroforestry Systems. 2019;93(4):1319-1336. doi:10.1007/s10457-018-0233-7.
2. Anand RK, Khare N, Dwivedi SV, Singh M, Umrao R. Studies on adoption of traditional agroforestry among the farmers in district Sonbhadra, Uttar Pradesh. The Indian Forester. 2016;142(2):144-150.
3. Debta H. Agroforestry: an effective potential tool for ensuring quality life for the Indian farmers: a review. Agricultural Reviews. 2025;46(4):531-542. doi:10.18805/ag.R-2658.
4. Deshmukh H, Dobriyal M, Tandel MB, Gunaga R, Sharma OP, Garde YA, et al. Development and standardization of an innovative scale for measuring the socio-economic status of agroforestry farmers in South Gujarat, India. Sustainability. 2023;15(3):2691. doi:10.3390/su15032691.
5. Kashyap P, Prusty AK, Panwar AS, Paramesh V, Natesan R, Shamim M, et al. Achieving food and livelihood security and enhancing profitability through an integrated farming system approach: a case study from Western Plains of Uttar Pradesh, India. Sustainability. 2022;14(11):6653. doi:10.3390/su14116653.
6. Khosla S, Timilsina RR, Jena PR, Rahut DB. Regenerative agriculture practices and multidimensional poverty in eastern rural India. Scientific Reports. 2025;15(1):36159. doi:10.1038/s41598-025-18461-5.
7. Naveen Y, Priya Rajendra M, Chiranjeeva Reddy M, Sarvade S, Gouthami G, Sri Ram M, et al. Evaluating agroforestry systems for sustainable land management and climate resilience in Siddipet District, central agroclimatic zone, Telangana, India. Agroecology and Sustainable Food Systems. 2025;49(8):1349-1367. doi:10.1080/21683565.2025.2508203.
8. Nayal R, Thomas M, Chalimalla M, Dhyani S. Participatory analysis of agroforestry systems: bridging ecosystem services and farmer resilience in drylands. Climate Change and Environmental Sustainability. 2024;12(1):53-64. doi:10.5958/2320-642X.2024.00006.7.
9. Palsaniya DR, Kumar S, Das MM, Rai SK, Kumar TK, Kumar S, et al. Rain water harvesting, agroforestry and goat based intensification for livelihood resilience in drought prone rainfed smallholder farming system: a case for semiarid tropics. Agroforestry Systems. 2023;97(8):1405-1419. doi:10.1007/s10457-023-00863-x.
10. Pande CV. Farm-Forestry, smallholder farms and policy support -the way ahead. In: Shukla G, Chakravarty S, Panwar P, Bhat JA, editors. Agroforestry-Small Landholder's Tool for Climate Change Resiliency and Mitigation. IntechOpen; 2021. doi:10.5772/intechopen.96942.
11. Raj A, Jhariya MK, Yadav DK, Banerjee A, Meena RS. Agroforestry: a holistic approach for agricultural sustainability. In: Jhariya MK, Banerjee A, Meena RS, Yadav DK, editors. Sustainable Agriculture, Forest and Environmental Management. Singapore: Springer; 2019. p. 101-131. doi:10.1007/978-981-13-6830-1_4.
12. Ram A, Dogra A, Dev I, Kumar N, Gautam K, Londhe S, et al. Agroforestry for sustainable cereal based cropping systems in Indo-Gangetic Plain Region. In: Kashyap PL, Gupta V, Gupta OP, Sendhil R, Gopalareddy K, Jasrotia P, et al., editors. New Horizons in Wheat and Barley Research. Singapore: Springer Nature; 2022. p. 617-637. doi:10.1007/978-981-16-4134-3_21.
13. Shukla HO, Tomar A, Kushwaha A, Singh R, Srivastav A. Agroforestry Practices in Ballia District of Eastern Plain Region of Uttar Pradesh, India. 2020. doi:10.5281/ZENODO.3633605.
14. Singh AK, Gohain I, Datta M. Upscaling of agroforestry homestead gardens for economic and livelihood security in mid-tropical plain zone of India. Agroforestry Systems. 2016;90(6):1103-1112. doi:10.1007/s10457-015-9886-7.
15. Srivastav A, Prakash S, Barman PK, Kumar D. Contribution of Agroforestry Tree Species in Vindhyan Region of Uttar Pradesh, India. 2024. doi:10.5281/ZENODO.14015940.
16. Verma P, Bijalwan A, Dobriyal MJR, Swamy SL, Thakur TK. A paradigm shift in agroforestry practices in Uttar Pradesh. Current Science. 2017;112(3):509-516. doi:10.18520/cs/v112/i03/509-516.
17. Yadav G, Jat HS, Raju R, Yadav RK, Singh SK, Chaudhari SK, et al. Enterprise mix diversification: an option for ecologically sustainable food and nutritional security of small holders in Indo-Gangetic plains. International Journal of Agricultural Sustainability. 2022;20(1):31-41. doi:10.1080/14735903.2021.1912978.
18. Yashmita-Ulman, Patel A. Agroforestry systems in Sultanpur district, Uttar Pradesh, India: a gender perspective. Range Management and Agroforestry. 2025;46(2):203-213. doi:10.59515/rma.2025.v46.i2.04.