Trends and dynamics of crop diversification in Nanded district of Maharashtra: a 25 year analysis

Introduction

Agricultural diversification refers to the reallocation of productive resources among crops, enterprises and value chains in response to technology, markets, natural resources and risk. At the crop level, diversification may involve replacing a dominant crop, adding crops, or distributing land more evenly across existing crops. These pathways have different implications. A portfolio with many crops can remain highly concentrated when most land is allocated to only one or two commodities; consequently, both crop richness and evenness must be examined (Simpson, 1949; Shannon, 1948) [1,2] . Diversification is especially relevant in rainfed and semi-arid agriculture. A broader crop portfolio can spread climatic, biological and price risk, improve temporal use of labour and land, support rotations and create access to higher-value markets. However, specialization can also be a rational response when one crop benefits from superior prices, input delivery, processing facilities or procurement. Empirical assessment should therefore avoid assuming that every shift away from food grains represents durable diversification (Joshi et al., 2006; Chand, 1996; Birthal et al., 2020) [3,4,5] . Indian agriculture has undergone marked compositional change with the expansion of oilseeds, pulses, horticulture and other commercial crops. Previous studies show that irrigation, infrastructure, market access, price incentives, technology and farm resources influence whether this process produces a balanced crop mix or concentration in a profitable commodity (Chand, 1996; Birthal et al., 2020; Kumar et al., 2012; Negi et al., 2020) [4,5,6,7] . District-level evidence is particularly important because agro-climatic conditions and market linkages vary substantially within states. Nanded district lies in the Marathwada region of Maharashtra and contains a large rainfed farming economy exposed to rainfall variability and commodity-market fluctuations. Its cropping pattern has changed markedly during the past quarter century, including expansion of soybean and gram and contraction of traditional cereals. Yet the long-run direction of diversification is not evident from crop areas alone. A fall in cereal area may initially widen the crop portfolio but later increase concentration if the released area is absorbed primarily by a single oilseed.

The study therefore aimed to quantify the extent, trend and structural dynamics of crop diversification in Nanded district from 2000-01 to 2024-25. It asked three questions: whether diversification increased over the entire period; whether conclusions were robust across complementary indices and statistical tests; and which crop groups and individual crops drove the observed movement. The paper contributes a transparent district-level assessment that separates crop substitution from broad-based diversification and connects index movements to extension and policy priorities.

Materials and methods

Study area data and period

The unit of analysis was Nanded district, Maharashtra. The study used annual secondary data for 25 agricultural years, 2000-01 to 2024-25, compiled from published district-level statistical records. The analytical file contained area under rice, wheat, kharif jowar, rabi jowar, bajra, maize, gram, mung, tur, udid, total groundnut, sunflower, safflower, soybean, sesamum, cotton, sugarcane, and fruits and vegetables. These 18 categories were mutually exclusive. Total groundnut was used to avoid double counting its seasonal components. Crop areas and total cropped area were expressed in hundred hectares.

A missing cotton-area observation for 2007-08 was reconstructed from the reported cotton production and productivity using the source units and the standard bale conversion. The reconstructed value was 2,078 hundred ha. No other observation was altered. Annual totals were recalculated from the mutually exclusive crop categories before the indices were estimated. The 25 years were also grouped into five consecutive five-year periods: I (2000-01 to 2004-05), II (2005-06 to 2009-10), III (2010-11 to 2014-15), IV (2015-16 to 2019-20) and V (2020-21 to 2024-25).

Diversification and concentration measures

For crop i in year t, the area share was 𝑝 𝑖𝑡 = 𝐴 𝑖𝑡 ∑ 𝑖 𝐴 𝑖𝑡 . The Herfindahl index (HI) was the sum of squared crop shares and the Simpson diversification index (SID) was 1 -𝐻𝐼. Shannon entropy was 𝐸 𝑡 = -∑ 𝑖 𝑝 𝑖𝑡 ln (𝑝 𝑖𝑡 ). Because the number of crops with positive area varied slightly, entropy was normalized as 𝑁𝐸𝐼 𝑡 = 𝐸 𝑡 ln ( 𝑁 𝑡 )

, where N t is the number of positive-area crops. The composite entropy index was 𝐶𝐸𝐼 𝑡 = 𝑁𝐸𝐼 𝑡 (1 -1

𝑁 𝑡

). The effective number of crops was 1 𝐻𝐼 , and CR3 was the combined share of the three largest crops. Higher SID, NEI, CEI and effective crop number indicate greater diversification; higher HI and CR3 indicate concentration (Simpson, 1949; Shannon, 1948; Herfindahl, 1950) [1,2,8] .

Trend structural change and instability analysis

Annual indices were summarized by arithmetic means within each five-year period. Crop-group shares were calculated for cereals, pulses, oilseeds, commercial crops, and fruits and vegetables.

Structural change at crop level was measured as the difference between mean shares in the terminal triennium (2022-23 to 2024-25) and base triennium (2000-01 to 2002-03), reducing sensitivity to a single endpoint year. Linear index trends were estimated against a time variable. Statistical inference used Newey-West heteroskedasticity-and-autocorrelationconsistent standard errors with lag one and finite-sample correction (Newey et al., 1987) [9] . Robustness was checked with the distribution-free Mann-Kendall trend test and Theil-Sen median slope (Mann, 1945; Sen, 1968) [10,11] . Crop-area instability was assessed with the Cuddy-Della Valle index, CDVI=CV√(1-R 2

), where R 2 came from a linear time trend (Cuddy et al., 1978) [12] . Results were reported to four decimal places throughout so numerical statements correspond exactly with tables.

Results and discussion

Long run extent of crop diversification

The annual indices provide a direct test of whether cultivated area became more evenly distributed. All complementary measures pointed to the same long-run outcome (Table 2 and Figure 1). SID declined from 0.8026 in 2000-01 to 0.7168 in 2024-25, while HI increased from 0.1974 to 0.2832. CEI fell from 0.6720 to 0.5550. The effective number of crops declined by 1.5353, from 5.0662 to 3.5309, whereas CR3 increased by 0.1233, from 0.6698 to 0.7931. Thus, by 2024-25 almost four-fifths of total crop area was concentrated in the three leading crops. The simultaneous fall in SID, CEI and effective crop number, together with the rise in HI and CR3, rules out an interpretation based on a peculiarity of one index. Nanded continued to report a large crop count, but area became progressively less even. This distinction is important: crop richness remained broadly present, whereas economic land allocation became concentrated. Comparable Indian evidence shows that diversification outcomes depend on the interaction of relative profitability, markets, irrigation and technology rather than on the mere availability of alternative crops (Chand, 1996 [4,5,6,7] .

Period wise dynamics

Five-year averages smooth weather-driven annual movements and reveal a clear two-stage path (Table 3). Diversification improved modestly between Period I and Period II: mean SID rose from 0.8360 to 0.8446 and CEI from 0.7205 to 0.7266. The effective crop number also increased from 6.1838 to 6.5268. This early gain was reversed thereafter. By Period V, SID had declined to 0.7439, CEI to 0.5775 and effective crop number to 3.9277, while CR3 had risen to 0.7751. The period evidence indicates that Nanded did not follow a uniform linear transition. Initial reallocation away from traditional cereals produced a somewhat more even portfolio, but subsequent expansion became concentrated in a narrower set of crops. This pattern is consistent with the proposition that market development can first facilitate diversification and later encourage regional specialization where one crop acquires a strong comparative advantage (Chand, 1996; Kumar et al., 2012) [4,6] .

Statistical evidence on the direction of change

The decline was statistically systematic rather than an artifact of endpoint selection (Table 4). CEI recorded a HAC slope of -0.0072 per year (p < 0.001) and a Theil-Sen slope of -0.0079. SID declined at -0.0050 per year (p < 0.001), while HI increased by the same magnitude because 𝑆𝐼𝐷 = 1 -𝐻𝐼. CR3 rose by 0.0078 per year (p < 0.001).

Mann-Kendall tests supported the same directions. The agreement be-tween parametric and rank-based methods is important because agricultural time series can contain heteroskedasticity, serial dependence and influential weather years. The effective number of crops fell by about 0.1437 per year in the linear specification. This metric translates concentration into an intuitive equivalent crop count: the observed distribution in the final year had the same concentration as only 3.5309 equally represented crops. The evidence therefore supports a conclusion of statistically significant reconcentration over the full study horizon.

Transformation of broad crop groups

The broad crop composition explains the shift in the indices (Table 5 and Figure 2). Cereals declined from 33.0900% of cropped area in Period I to 5.4226% in Period V. Oilseeds increased from 11.0022% to 44.6281%, while pulses rose from 23.1984% to 28.4396%. Commercial crops accounted for 31.3380% in Period I, peaked at 35.6672% in Period III, and declined to 20.0650% in Period V. Fruits and vegetables remained below 2% in every period. The early reduction in cereal dominance contributed to diversification, but the later oilseed expansion was overwhelmingly soybeanled. Hence, movement away from cereals should not automatically be interpreted as continuing diversification. Diversification toward highvalue commodities can support agricultural growth and income (Joshi et al., 2006; Birthal et al., 2020) [3,5] , but durable resilience requires several viable alternatives and supporting value chains rather than dependence on a new dominant commodity.

Crop level drivers of diversification and concentration

Base-to-terminal triennium comparison identifies the crops responsible for reconcentration (Table 6 and Figure 3). Soybean gained 42.1502 percentage points and moved from rank 10 to rank 1. Gram gained 15.3798 points and rose from rank 8 to rank 2. Sugarcane, tur, maize, fruits and vegetables, and groundnut registered smaller gains. These increases were offset chiefly by kharif jowar (-23.4277 points), cotton (-16.2734), udid (-7.4298), rice (-3.0842), mung (-2.9781) and wheat (-2.6771).

The crop portfolio therefore underwent substitution rather than a uniform expansion of alternatives. Soybean and gram together absorbed much of the area released from kharif jowar, cotton and minor crops. The soybean area trend was 207.2997 hundred ha per year, with endpoint CAGR of 18.1098%, while gram also displayed strong expansion. Although this transition may reflect perceived profitability, production suitability and market access, soybean dominance exposes farm households and the district economy to correlated rainfall, pest and price risks. Extension advice should consequently focus on economically coherent rotations and enterprise combinations, not indiscriminate addition of crops.

Implications for agricultural extension and development

The results alter the practical interpretation of diversification in Nanded. The central challenge is not the absence of multiple crops but the declining evenness of their area distribution. Extension programmes should demonstrate locally suitable soybean-pulse-cereal and soybean-oilseed rotations, incorporate farm-budget and risk comparisons, and link recommendations to seed availability, mechanization, storage, aggregation and assured market channels. Traditional and climate-resilient cereals require attention where they remain agronomically and nutritionally suitable.

Policy should monitor both profitability and concentration. District planning systems can report SID or CEI alongside crop area, yield and value of output. A diversification programme should be judged by whether several alternatives acquire viable shares and stable value chains, rather than whether a minor crop was introduced. Farmer producer organisations, market intelligence, pulse and oilseed processing, water management and weather-based advisories can lower transition costs and help farmers choose portfolios suited to farm resources and risk preferences.

Conclusion

Crop diversification in Nanded district improved only during the early part of the study and then reversed. Over 2000-01 to 2024-25, SID and CEI declined significantly, effective crop number contracted, and the share of the three leading crops increased. The decisive structural change was the rise of soybean, accompanied by gram, against a sharp decline in kharif jowar, cotton and several minor crops. Thus, the district shifted crops but did not achieve sustained broad-based diversification; it moved toward soybean-led concentration. The policy priority is to develop profitable, agronomically compatible and market-supported crop baskets that preserve farm income opportunities while reducing dependence on a small number of commodities.

References

1. Simpson EH. Measurement of diversity. Nature. 1949;163:688.
2. Shannon CE. A mathematical theory of communication. Bell System Technical Journal. 1948;27:379-423, 623-656.
3. Joshi PK, Birthal PS, Minot N. Sources of agricultural growth in India: Role of diversification towards high-value crops. MTID Discussion Paper No. 98. Washington DC: International Food Policy Research Institute; 2006.
4. Chand R. Diversification through high-value crops in western Himalayan region: Evidence from Himachal Pradesh. Indian Journal of Agricultural Economics. 1996;51(4):652-663.
5. Birthal PS, Hazrana J, Negi DS. Diversification in Indian agriculture towards high-value crops: Multilevel determinants and policy implications. Land Use Policy. 2020;91:104427.
6. Kumar A, Kumar P. Crop diversification in Eastern India: Status and determinants. Indian Journal of Agricultural Economics. 2012;67(4):600-616.
7. Negi DS, Birthal PS, Roy D, Khan MT. Market access, price policy and diversification in Indian agriculture. IGIDR Working Paper WP-2020-009. Mumbai: Indira Gandhi Institute of Development Research; 2020.
8. Herfindahl OC. Concentration in the U.S. steel industry. Doctoral dissertation. New York: Columbia University; 1950.
9. Newey WK, West KD. A simple, positive semi-definite, heteroskedasticity and autocorrelation consistent covariance matrix. Econometrica. 1987;55(3):703-708.
10. Mann HB. Nonparametric tests against trend. Econometrica. 1945;13(3):245-259.
11. Sen PK. Estimates of the regression coefficient based on Kendall's tau. Journal of the American Statistical Association. 1968;63(324):1379-1389.
12. Cuddy JDA, Della Valle PA. Measuring the instability of time series data. Oxford Bulletin of Economics and Statistics. 1978;40(1):79-85.
13. Government of Maharashtra. Agricultural statistics of Maharashtra, relevant annual issues. Pune: Department of Agriculture. How to cite this article: Kadam RB, Kamble SH, Shelke RD, Bokan SD, Jadhav EP. Trends and dynamics of crop diversification in Nanded district of Maharashtra: a 25 year analysis. International Journal of Agriculture Extension and Social Development. 2026;9(9):813-818. doi: https://www.doi.org/10.33 545/26180723.2026.v9.i9k.4273