The study using cross-sectional data collected from 300 dairy farmers has analyzed the factors affecting adoption of improved forage technologies and its impact on milk yield and feed sufficiency in central region. We used inverse-propensity-weighting regression adjustment (IPWRA) method as main technique for impact evaluation and checked the robustness of the results using matching methods. Our findings suggest that education status, adult cattle unit, animal breed type, off-farm income activities, farm size, and access to training and market significantly influence adoption of improved forage technologies and practices. Further, the adoption led to a significant increase in daily milk yield (1.07 to 1.34 L), total dry matter availability by over 27%, and green fodder availability by around 80%. Ration balancing has been identified as a significant concern in the study region. Consequently, the study suggests that adopting a comprehensive approach is necessary to address the issue of proper ration balancing and fully harness the production potential of dairy animals.

1 INTRODUCTION

The dairy industry worldwide is becoming increasingly intensified in response to growing human populations, reduced land availability per capita, and higher demand for animal-based food products (Adegbola & Dahl, 2020). This sector plays a vital role in ensuring food security, particularly in developing countries like India, and offers a promising means of livelihood for impoverished households (Choudhary & Sirohi, 2022; Saxena et al., 2019). Not only does dairying promote equitable income distribution among farming households (Choudhary & Singh, 2019) but it also serves as a reliable safeguard against natural disasters like droughts, famines, and other calamities. Hence, it is imperative to prioritize the advancement of this sector in order to achieve sustainable development goals.

India possesses the largest bovine population globally; however, the productivity potential of Indian animals remains significantly low, thereby compromising the profitability of dairy farmers (Choudhary & Sirohi, 2022). A major contributing factor to this issue is the scarcity of green fodder, particularly during the dry season, which has been widely acknowledged (Ghosh et al., 2016). In India, the allocation of land for cultivating green fodder crops is limited and has rarely exceeded 5% of the total cultivated area. Consequently, the availability of feed and fodder has consistently fallen short of the required norms, thereby restricting the realization of the livestock's true production potential (Sharma et al., 2021). Currently, the country faces deficits of 11.24% in green fodder, 23.4% in dry crop residues, and 28.9% in concentrate feed ingredients (Roy et al., 2019). Moreover, there are regional and seasonal disparities in fodder production and availability, with most deficient areas located in arid and semiarid regions exposed to challenging climates. Therefore, addressing the gaps in feed and fodder supply in vulnerable landscapes is of particular importance for sustaining the livelihoods of smallholder dairy farmers. Moreover, enhancing the forage resource base through improved cultivation practices and resource management is imperative for bolstering the productivity and resilience of India's dairy sector and for ensuring its continued contribution to the global milk supply chain.

The potential for enhancing milk productivity and profitability is also contingent upon the feeding management by dairy farmers. Adequate feed, usually measured as dry matter intake (DMI), ensures that animals meet their nutrient requirements, obtain sufficient energy, maintain gut health, support protein synthesis, enhance immune function, and achieve optimal production and performance (Leoni et al., 2022). Several experimental field trials have shown the potential of integrating improved forages in enhancing livestock productivity (Ghosh et al., 2016; Sharma et al., 2007); yet, comprehensive and quantitative evidences regarding on-farm impact of adoption of improved forage technologies on DMI and milk productivity of livestock are lacking in the Indian context.

The present study, therefore, using the example of the KISAN MITrA (Knowledge-based Integrated Sustainable Agriculture Network Mission India for Transforming Agriculture) project seeks to fill the literature gaps on the on-farm impacts of forage-based interventions on milk yield and feed sufficiency. The project aimed to provide technological assistance to as many farmers as possible, adopting an open participation policy. The focus was on small-scale livestock farmers, common in the region, so no specific selection criteria were needed. The project aimed to improve forage production, conservation, and utilization techniques, offering support at no cost. Under the project, a broad set of improved green forage technologies and practices like use of quality fodder seeds of berseem (Trifolium alexandrinum), oat (Avena sativa), and sorghum (Sorghum bicolor); Napier grasses on bunds and blocks; and on-farm demonstrations on cultural practices of fodder production, conservation, and utilization were promoted (Annexure 1) making it an ideal context for an investigation of the aforementioned farm-level adoptions and impact analysis of forage-based interventions. The study seeks to provide comprehensive and quantitative evidence essential for informed decision-making by dairy farmers and thus would contribute to the development of sustainable and profitable dairy farming practices in semiarid regions across the globe.

2 MATERIAL AND METHODS

2.1 Study area, sampling, and data

This study was conducted in three districts, namely, Jhansi, Lalitpur, and Jalaun located in Bundelkhand region of central India (Annexure 2). The region is characterized by arid and semiarid climatic conditions, posing significant challenges to agricultural productivity and rural livelihoods (Choudhary et al., 2024). In response to climate-related risks to crop production, livestock-based mixed farming has emerged as a crucial component of the local economy, providing a sustainable livelihood option for the rural communities of the region. The study site represents typical traditional smallholder dairy farming regions in the semiarid tropics of India (Sharma et al., 2021).

The study used purposive and random sampling techniques to select participants. Initially, three villages per district were purposefully chosen as “treated villages,” where project activities were concentrated. Furthermore, neighboring villages without forage-based interventions but with similar agro-climatic, infrastructural, and socio-economic characteristics were randomly selected as “control villages.” In the second stage, dairy households were randomly chosen within these villages. A total of 300 respondents were surveyed, comprising 150 from treated villages and 150 from control villages. The survey took place over two visits, spanning from June 2021 to March 2022. Each farmer was visited twice to collect data on their current production and feeding practices, minimizing recall bias.

2.2 Analytical framework

2.2.1 Dry matter requirement, availability, and balance

The total forage requirement for the livestock population in the region was estimated based on the types of livestock owned by farmers. Livestock numbers were converted to adult cattle units (ACUs) using conversion factors developed by Kumbhare et al. (1983). Crop residues, such as sorghum, maize, paddy, wheat, coarse cereals, pulses, and groundnut, were considered as a source of dry fodder. The production of dry matter from green fodder was estimated by considering the acreage of fodder crops and their respective dry matter content. The availability and requirement of concentrate on a dry matter basis were estimated using the production of cereals, oilseeds, and pulses, accounting for 10% of total production as broken discarded grains. The estimated dry matter availability was calculated by combining data from crops, green fodder, free range grazing, and available concentrates. The deficit or surplus between dry matter availability and requirement was determined. Annapratha, the practice of “free range grazing,” is commonly followed during the lean and rainy seasons (Choudhary, Sharma, et al., 2022; Saran et al., 2000). The dry matter consumed through Free Range Grazing in the Bundelkhand region was estimated based on Dwivedi and Singh (2012). The dry matter requirement for maintenance of an ACU was estimated at 2.5% of their body weight. According to ration balancing guidelines, 2/3 of this requirement should come from roughages and 1/3 from concentrate. The roughages part is further divided into 2/3 as dry fodder and 1/3 as green fodder, as per standard requirements (ICAR, 2001).

2.2.2 Impact evaluation

Impact can be measured by calculating the average treatment effect on the treated (ATT) as follows.

ATT=E\left\{{Y}_{iA}-{Y}_{iNA}|{T}_i=1\right\}

(1)

where E{.} is the expectation operator.

{Y}_{iA}

indicates potential outcome for ith adopter household (

{T}_i=1\Big)

, and

{Y}_{iNA}

is potential outcome for ith nonadopter household (

{T}_i=0\Big)

The key estimation problem in the above equation (1) is that it is not possible to observe the outcome for adopters had they not adopted, that is, E( ${Y}_{iNA}\left|{T}_i=1\right).$ Takahashi and Barrett (2013) further opined that replacing these unobserved counterfactuals by outcomes of nonadopters may result in biased ATT estimates.

Therefore, we used inverse-propensity-weighting regression adjustment (IPWRA) method—a double robust estimator of ATT (Wooldridge, 2010), as our prime estimator. IPWRA uses inverse probabilities of treatment as weights to compute regression coefficients that are further used for estimating treatment effects (ATT).

For constructing IPWRA estimator, we first specified the outcome equation as

{Y}_i={\alpha}_i+{\sum}_{j=1}^n{\beta}_{ij}{X}_{ij}+{\varepsilon}_i

(where i = 1 for adopters and 0 for nonadopters; X is as defined in equation 1) and estimated propensity scores, that is, the probability of receiving treatment

P\left(X,\hat{Y}\right)

using probit model. The ATT was then computed as

{ATT}_{IPWRA}={n}_A^{-1}\sum \limits_{i=1}^n{T}_i\ \left[{r}_A\left(X,{\delta}_A\right)-{r}_{NA}\left(X,{\delta}_{NA}\right)\right]

(2)

where n_A is the number of adopters (A), r_A(X) and r_NA(X) are the postulated regression model for the adopters (A) and nonadopters (N) based on observed covariates X, respectively. The parameter

{\delta}_A

= (

{\alpha}_A,{\beta}_A\Big)

in the above equation (2) is obtained using inverse-probability-weighted least squares procedures as follows.

\underset{\alpha_A{\beta}_A}{\min}\sum \limits_{i=1}^N{T}_i{\left({y}_i-{\alpha}_A-X{\beta}_A\right)}^2/\hat{p}\left(X,\hat{\gamma}\right)

(3)

Similarly,

{\delta}_{NA}

= (

{\alpha}_{NA},{\beta}_{NA}\Big)

is computed as:

\underset{\alpha_{NA}{\beta}_{NA}}{\min}\sum \limits_{i=1}^N\left(1-{T}_i\right){\left({y}_i-{\alpha}_{NA}-X{\beta}_{NA}\right)}^2/\left(1-\hat{p}\left(X,\hat{\gamma}\right)\right)

(4)

Imbens and Wooldridge (2009) advocated for employing multiple methodologies to estimate treatment effects, ensuring the robustness of findings, given the availability of diverse techniques. As a pivotal robustness check, our study also incorporated propensity score matching (PSM) and coarsened exact matching (CEM), the two widely utilized techniques in impact evaluation that offer a potent means of addressing missing data by matching individuals according to their propensity scores.

3 RESULTS

3.1 Choice of variables and its descriptive statistics

The household serves as the primary beneficiary of farm technology, making household characteristics such as household size, education status, and farming experience crucial considerations in the adoption process. Additionally, farm characteristics and institutional factors have been identified as significant influencers in the technology adoption process (Sharma et al., 2021). The dissimilarities in observed characteristics between households from treated and control villages are clearly apparent (Table 1).

TABLE 1. Definitions and sample averages of selected variables.

Variables	Description	Control (C, n = 150)	Treated (T, n = 150)	Mean difference (C − T)
Households characteristics
Age	Age of household head (years)	45.01	46.62	−1.61
Experience	Experience of household head in dairy farming (years)	25.41	26.83	−1.42
Education	Numbers of years of schooling by household head	3.99	5.27	−1.28*
HH_Size	Household size (no.)	5.26	6.18	−0.92*
Dependency ratio	(household members <15 and >65 years)/household size	0.286	0.289	0.003
Farm characteristics
Land holdings	Operational holding in hectares	1.52	1.91	−0.39**
ACU	Adult cattle unit	5.05	6.11	−1.06*
Buffalo to cattle ratio	Buffalo to indigenous cattle ratio in dairy herd	0.43	0.57	−0.14*
Off-farm	% of household involved in off-farm income activities (%)	36.18	41.13	−4.95*
Institutional characteristics
Training	% of households heads exposed to training and demonstration visit	63.29	91.58	−28.29*
Credit	% of households that has access to farm credit	46.22	47.57	−1.35
Market	% of households that are able to sale surplus milk	42.16	63.47	−21.31*
Outcome variable
Milk yield	Milk yield in liter per day	3.23	4.27	−1.04*
DM_Dry fodder	Dry matter supplied as percentage of required dry fodder	65.51	86.00	−20.48*
DM_Green fodder	Dry matter supplied as percentage of required green fodder	73.31	154.60	−81.28*
DM_Concentrate	Dry matter supplied as percentage of required concentrate	6.14	7.71	−1.57*
DM_Total feed	Dry matter supplied as percentage of total feed requirement	112.59	140.29	−27.69*

* p < 0.01, and
** p < 0.05.

Specifically, compared to control villages, household heads in treated villages exhibit higher levels of education and possess larger land holdings. Adopters, on average, have a higher cattle unit (ACU) compared to nonadopters. Larger proportions of households in the treated village (41.13%) derive income from off-farm sources. Further, adopters are better exposed to training and demonstrations. Consequently, a significant proportion of adopters (63.47%) are able to sell surplus milk. The significant differences in outcome indicators clearly indicate that adopters of improved forage technology are systematically better off than their nonadopters counterpart in terms of milk yield and feed availability for livestock (on dry matter basis) (Table 1). However, caution must be exercised in drawing causal claims about the impact of forage-based interventions on these indicators, as confounding factors have not been accounted for.

3.2 Drivers of the adoption of forage technologies

We have previously motioned that IPWRA estimator for ATT requires estimation of propensity scores. The present study employed probit model to estimate propensity scores for our IPWRA estimator. The marginal effects of this model can be found in Table 2. It is important to note that propensity score estimation is solely employed to achieve covariate balance between adopters and nonadopters. Visual observation of the Figure 1 clearly indicate that there is considerable overlap of the distributions of the propensity scores for adopters and nonadopters of improved forage technology after matching suggesting that the assumption of common support firmly holds.

TABLE 2. Regression results from probit model on matched sample^a for the driving factors of adoption of improved forage technologies.

Variables	Coefficients	Std. error	Marginal effect
Age	0.049	0.061	0.0017
Experience	−0.021	0.032	−0.004
Education	0. 218*	0.007	0.032
HH_Size	0.035	0.013	0.017
Dependency ratio	−0.114*	0.010	−0.013
Land holdings	0.273*	0.001	0.031
ACU	0.349*	0.011	0.146
Buffalo to cattle ratio	0.417*	0.131	0.117
Off-farm	0.027**	0.011	0.032
Training	0.468*	0.107	0.034
Credit	0.016	0.022	0.006
Market	0.226*	0.022	0.027
Log likelihood	−336.17
LR χ²	59.36*
Sample size^b	276

Abbreviation: LR, likelihood ratio.
a We also used NMM and calliper-matching methods, and the results were similar to KBM based estimates. Hence, in the interest of time and space, we present results for KBM method only.
b KBM resulted into 141 and 135 observations from treated and control samples, respectively, making total sample size of 276.
* p < 0.01, and
** p < 0.05.

Details are in the caption following the image — **FIGURE 1**
Open in figure viewer PowerPoint

Propensity score distribution of treated and controls showing the common support.

The findings presented in Table 2 indicate that the probability of the adoption of forage technologies significantly increased education of the family head. Households with larger farm sizes demonstrate a higher likelihood of adopting improved forage technologies. Additionally, households with a higher proportion of economically inactive members are less likely to adopt forage-based interventions. As per our expectation, households possessing a greater number of livestock units and a larger herd of buffaloes are also more inclined to adopt improved forage interventions, with adoption probabilities increasing by 0.146 and 0.117, respectively. Furthermore, a positive and significant correlation was identified between off-farm income and the adoption likelihood of improved forage technologies. Moreover institutional factors, specifically training and market access, play a significant and positive role in the adoption of forage technologies.

3.3 Estimation of causal impact

ATT results in IPWRA indicate that adoption of improved forage cultivation practices helped in increasing daily milk productivity of dairy animals by 1.34 L per day (Table 3). From the results of different approaches, it can be stated that an increase in productivity could be between 1.07 and 1.34 L per day. The assessment of the impact of adoption on feed sufficiency revealed that the project interventions in the treated villages significantly increased the availability of green fodder (on dry matter basis) by over 84% and total dry matter (as a percentage of the requirement) by over 27% compared to the control villages. It is important to highlight that this increase in availability was primarily attributed to the rise in green fodder availability in the treated villages, which aligns with the main objective of the KISAN MITrA project.

TABLE 3. Treatment effect on outcome variables.

Methods	Treatment effects	Daily milk yield		DM_Dry fodder		DM_Green fodder		DM_Concentrate		Dry matter_Total
Methods	Treatment effects	Estimate	Robust SE	Estimate	Robust SE	Estimate	Robust SE	Estimate	Robust SE	Estimate	Robust SE
IPWRA	ATE	1.07*	0.15	21.14	17.12	81.28*	10.22	1.57	1.59	27.69*	7.13
IPWRA	ATT	1.34*	0.23	16.91	11.21	86.30*	13.02	1.33	0.98	26.81*	4.41
PSM	ATE	1.11*	0.13	19.32	16.73	78.64**	39.71	1.44	1.15	28.54*	6.32
PSM	ATT	1.25*	0.20	16.59	11.52	85.17**	43.03	1.13	1.08	27.10*	5.19
CEM	ATE	1.07*	0.11	20.14	15.09	80.18*	8.56	1.74	1.13	29.62*	7.59
CEM	ATT	1.34*	0.19	15.98	11.46	84.61*	11.24	1.21	1.04	27.87*	4.74

Abbreviations: ATE, average treatment effect; ATT, average treatment effect on the treated; CEM, coarsened exact matching; DM, dry matter; IPWRA, inverse-propensity-weighting regression adjustment; PSM, propensity score matching.
* p < 0.01, and
** p < 0.05.

We did not find significant impact on the availability of dry fodder as well as concentrate. The percentage of dry fodder and concentrate supplied (on a dry matter basis) indicated severe shortage in all villages, although the deficit was slightly less in the treated villages.

4 DISCUSSION

The present study sheds light on the intricate dynamics of technology adoption in agriculture, emphasizing the role of household and farm-specific characteristics. These results suggest that educated households with more number of economically active members possess a greater ability to comprehend new information and assess the significance of such technologies when it comes to their adoption. The direct relation between inclination towards adoption of improved agricultural technologies and education of household head has good literature support (Sharma et al., 2021).

Larger farm sizes increase the likelihood of adoption (Kabir & Rainis, 2015), and this can be attributed to the fact that larger landholdings provide households with greater capacity to take risks, thereby increasing their probability of embracing new technologies (Singh et al., 2023). The positive correlation between livestock units and buffalo herd size with adoption probability supports findings by Sharma et al. (2021) who also reported that farm households in central India region with higher livestock units and having more buffaloes in their herd are more likely to adopt improved forage-based interventions—the probability increases by 0.099 and 0.132, respectively. Kanyenji et al. (2020) also highlighted the relationship between higher livestock units and increased adoption of agricultural technologies promoting forage yields. Additionally, Kassie et al. (2018) reported that owning productive breeds of livestock results in increased demand for feed. Notably, off-farm income emerges as a significant factor, alleviating liquidity constraints and enabling monetary investments in technology adoption. The positive correlation between off-farm income and technology adoption is consistent with existing literature (Choudhary & Singh, 2019; Diiro, 2013). The greater adoption intensity of agricultural technology among households with off-farm income, compared to those without such income, is widely acknowledged (Mwangi & Kariuki, 2015). The role of institutional factors in promoting adoption of agricultural technologies is well established (Choudhary, Dev, et al., 2022; Sharma et al., 2021). Exposure to trainings helps in enhancing farmers' confidence and mitigating the impact of lack of formal education on decision-making. Maina et al. (2020), in their study from Kenya, established that extension support to dairy farmers increases the probability of adoption of forage grass by 0.504. Similarly, in another study from the Bundelkhand region of India (Sharma et al., 2021), exposure to training and demonstrations on improved forage practices was found to increase the probability of adoption by 0.23. Additionally, access to output markets significantly influences the adoption and utilization of agricultural technologies among farmers (Saxena et al., 2019). In dairy production systems, farmers are more motivated to invest in and adopt technologies that enhance livestock production efficiency and quality when they have reliable and profitable channels to sell their milk (Kumar et al., 2022; Sharma et al., 2021). In the current study, we observed that the absence of efficient milk collection centers in the region has led to a prevalence of milk vendors or middlemen who dominate the milk marketing system. Consequently, farmers often receive inadequate prices for their milk. Nevertheless, some farmers manage surplus milk by processing it into higher-value dairy products like Ghee (clarified butter) and Khoa (highly condensed milk), which command better prices despite involving significant labor and processing costs. Hence, there is an urgent policy need in the region to develop a supportive market environment for surplus milk, encompassing both backward and forward market linkages. Promoting farmer producer organizations (FPOs) could be crucial in this regard to protect the interests of small dairy farmers.

The impact estimates provide compelling evidence of the positive impact of improved forage cultivation practices on daily milk productivity of dairy animals. The estimated increase of 1.34 L per day is substantiated by various approaches, suggesting a range between 1.07 and 1.34 L per day. This not only signifies a notable enhancement in productivity but also underscores the effectiveness of the project interventions. In a recent study from central India, Sharma et al. (2021) also reported that improved forage-based interventions translate into an increase in daily milk productivity of farm animals by more than 1 L. Moreover, the assessment of the impact on feed sufficiency reveals a substantial improvement in the availability of green fodder and total dry matter in the treated villages, with increases of 84% and 27%, respectively, compared to control villages. This aligns closely with the project's primary objective of augmenting green fodder availability. Introducing advanced practices for forage production, such as improved varieties of cereals and legume fodder crops, as well as perennial grasses like Bajra-Napier hybrid grass, agronomic managements (line sowing of seeds, integrated weed and pest management strategies, soil test-based nutrient application, and irrigation at critical crop stages), conservation (proper timing of harvest and use of silage additives to maintain forage quality during storage), and utilization (ration balancing with different types of fodder and integrating supplementary feeding with stored forages), played a crucial role in enhancing green fodder availability and bridging the demand–supply gap of forages in the treated villages. Furthermore, training and capacity building on integrating fodder crops into the existing cropping system, suitable fodder crops and their scientific package and practices including variety selection, planting management, integrated nutrient–weed–pest management etc., fodder conservation techniques, utilization of fodder in dairy nutrition, integrated farm management, and economic considerations in forage production significantly bolstered the knowledge and confidence of dairy farmers in taking up fodder production at their farms, consequently boosting milk productivity among their livestock.

The variations in the impact magnitudes observed across the three estimators (IPWRA, PSM, and CEM) may be attributed to bias resulting from unobserved factors (Wooldridge, 2010). These factors can lead to both underestimation and overestimation of treatment effects when utilizing matching techniques. It is important to note that the results obtained from the inverse probability weighting estimator with regression adjustment are only applicable to a subset of households that exhibit a more balanced distribution of independent variables. Considering the diverse biophysical, institutional, and economic contexts faced by different groups of farmers, it is plausible that the decision of many resource-poor farmers to not adopt technically beneficial technologies is actually an optimal choice. This notion is supported by previous studies highlighting that the supposed benefits of these technologies may not align with the realities experienced by the majority of small-scale farmers (Choudhary, Sharma, et al., 2022; Sharma et al., 2021). Consequently, the presence of unobserved heterogeneity among smallholders could explain why many farmers appear to avoid technologies that are promoted for their perceived advantages in developing countries, including India.

It is noteworthy that the impact on the availability of dry fodder and concentrate was not statistically significant, and all villages exhibited a considerable deficit, albeit slightly less in the treated villages. The results show the prevalence of imbalanced feeding practices in the region and are in congruence with Pathak et al. (2005), who also observed that the disproportionate supply of nutrients to livestock in central India is a major factor limiting livestock productivity. This underscores the need for targeted interventions to address the persistent shortage of dry fodder and concentrate, potentially through further project refinements or complementary measures. Overall, these research outcomes provide valuable insights for policymakers and practitioners in the agriculture sector, emphasizing the efficacy of specific interventions in enhancing milk productivity and the importance of addressing specific feed deficits for sustainable dairy farming practices.

In conclusion, the present study investigated the drivers as well as farm-level impacts of adoption of improved forage technologies promoted under KISAN MITrA project in Bundelkhand region of central India. The evidence of a significant and positive impact between project activities in treated villages and, milk productivity and feed sufficiency suggest that intensifying the dissemination of improved fodder based activities would be vital in changing livelihood of dairy farmers of semiarid regions. Ration balancing has emerged as a significant concern in the study region, emphasizing that simply increasing the availability of green fodder does not guarantee balanced feeding practices. Therefore, the study suggests that a comprehensive approach is required to address the issue of proper ration balancing for exploiting the full production potential of dairy animals.

The key policy variables influencing the adoption of improved forages are education, training, and off-farm income activities and market access. Therefore, mainstreaming practically oriented, participatory, and interactive model like farmer field school (FFS) program and encouraging frontline demonstrations by local research institutes to impart training to the dairy farmers on improved fodder production, conservation, and utilization would be imperative to improve farmers' capacity and skills in forage as well as dairy management. Additionally, fostering a supportive market ecosystem for surplus milk, including both upstream and downstream market connections, and establishing of Fodder Producer Organizations could play a pivotal role in scaling up the adoption of forage technology and maximizing the productivity potential of farm animals.

The evidence from the Bundelkhand region, known for its distinctive agro-ecological conditions, including undulating topography and unique climatic challenges, provides valuable insights into the promotion of improved forage technologies to enhance livestock productivity in arid and semiarid regions worldwide that encounter similar challenges. Nevertheless, it is crucial to integrate farmers' preferences and choices with the recommended policy interventions.

CONFLICT OF INTEREST STATEMENT

The authors declare no conflicts of interest.

ANNEXURE: PROJECT ACTIVITIES UNDER KISAN MITRA PROJECT

Forage crop production based intervention

Forage crops	Variety	Seed distribution (quintal)	On farm demonstration (number)
Kharif season (June–October)
Sorghum (Sorghum bicolour)	CSH-24, PC-6	32.0	264
Bajra–Napier hybrid grass (Pennisetum glaucum × P. purpureum)	IGFRI-6	62,000 (root slips)	149
Cowpea (Vigna unguiculat)	Kohinoor	5.0	132
Guar (Cyamopsis tetragonoloba)	BG-1,BG-2	5.0	131
Rabi (November–April)
Oat (Avena sativa)	Kent	29.0	231
Berseem (Trifolium alexandrinum)	Wardan	12.5	383
Zaid (may–June)
Sorghum (sorghum bicolour)	CSH-24, PC-6	10	107

Forage conservation and utilization based interventions

Intervention	Number
Silage preparation	15 demonstrations in each treated village
Hay preparation	15 demonstrations in each treated village
Training on balance forage feeding	25
Training on forage seed production	15 in each treated village

ANNEXURE: THE LOCALE OF THE STUDY AREA DELINEATING TREATED AND CONTROL VILLAGES

REFERENCES

Adegbola, T. A., & Dahl, G. E. (2020). Milk symposium introduction: Dairy production in developing countries. Journal of Dairy Science, 103, 9677–9680. https://doi.org/10.3168/jds.2020-18313
10.3168/jds.2020-18313
PubMed Google Scholar
Choudhary, B. B., Dev, I., Singh, P., Singh, R., Sharma, P., Chand, K., Garg, K. K., Anantha, K. H., Akuraju, V., Dixit, S., Kumar, S., Ram, A., & Kumar, N. (2022). Impact of soil and water conservation measures on farm productivity and income in the semi-arid tropics of Bundelkhand, central India. Environmental Conservation, 49, 263–271. https://doi.org/10.1017/S0376892922000352
10.1017/S0376892922000352
Google Scholar
Choudhary, B. B., Sharma, P., Singh, P., Kumar, S., Gupta, G., Kantwa, S. R., Upadhyay, D., Wasnik, V. K., Prasad, M., & Sharma, R. K. (2022). Green fodder cultivation improves technical efficiency of dairy farmers in semi-arid tropics of central India: A micro-analysis. Journal of Dairy Research, 89, 367–374. https://doi.org/10.1017/S0022029922000759
10.1017/S0022029922000759
CAS PubMed Google Scholar
Choudhary, B. B., & Singh, P. (2019). How unequal is rural Punjab? Empirical evidence from spatial income distribution. Current Science, 117, 1855–1862. https://doi.org/10.18520/cs/v117/i11/1855-1862
10.18520/cs/v117/i11/1855-1862
Google Scholar
Choudhary, B. B., Singh, P., Dev, I., Sharma, P., Singh, R., & Chand, K. (2024). Determinants of adoption of multiple natural resource management practices: A case study from semi-arid tropics of Central India. Environment Development and Sustainability. https://doi.org/10.1007/s10668-024-04995-7
10.1007/s10668-024-04995-7
Google Scholar
Choudhary, B. B., & Sirohi, S. (2022). Economic losses in dairy farms due to heat stress in sub-tropics: Evidence from North Indian plains. Journal of Dairy Research, 89, 141–147. https://doi.org/10.1017/S0022029922000371
10.1017/S0022029922000371
CAS Google Scholar
Diiro, G. (2013). Impact of off-farm income on technology adoption intensity and productivity: evidence from rural maize farmers in Uganda. International Food Policy Research Institute, Working Paper 11.
Google Scholar
Dwivedi, P. N., & Singh, H. (2012). Nutrient availability and milk production of cows in Bundelkhand region. Indian Veterinary Journal, 89, 40–42.
CAS Google Scholar
Ghosh, P. K., Palsaniya, D. R., & Srinivasan, R. (2016). Forage research in India: Issues and strategies. Agricultural Research Journal, 53, 1–12. https://doi.org/10.5958/2395-146X.2016.00001.6
10.5958/2395-146X.2016.00001.6
Google Scholar
ICAR. (2001). Nutrient requirements of cattle and buffalo. Indian Council of Agricultural Research, PUSA.
Google Scholar
Imbens, G. W., & Wooldridge, J. M. (2009). Recent developments in the econometrics of program evaluation. Journal of Economic Literature, 47, 5–86. https://doi.org/10.1257/jel.47.1.5
10.1257/jel.47.1.5
Web of Science® Google Scholar
Kabir, M. H., & Rainis, R. (2015). Adoption and intensity of integrated pest management (IPM) vegetable farming in Bangladesh: An approach to sustainable agricultural development. Environment, Development and Sustainability, 17, 1413–1429. https://doi.org/10.1007/s10668-014-9613-y
10.1007/s10668-014-9613-y
Web of Science® Google Scholar
Kanyenji, G. M., Oluoch-Kosura, W., Onyango, C. M., & Ng'ang'a, S. K. (2020). Prospects and constraints in smallholder farmers' adoption of multiple soil carbon enhancing practices in Western Kenya. Heliyon, 6, e03226. https://doi.org/10.1016/j.heliyon.2020.e03226
10.1016/j.heliyon.2020.e03226
CAS PubMed Google Scholar
Kassie, M., Stage, J., Diiro, G., Muriithi, B., Muricho, G., Ledermann, S. T., & Zeyaur, K. (2018). Push-pull farming system in Kenya: Implications for economic and social welfare. Land Use Policy, 77, 186–198. https://doi.org/10.1016/j.landusepol.2018.05.041
10.1016/j.landusepol.2018.05.041
Web of Science® Google Scholar
Kumar, S., Sharma, P., Satyapriya, G. P., Singh, M., Kumar, S., Halli, H. M., Choudhary, B. B., & Bagavathiannan, M. (2022). Economic impression of on-farm research for sustainable crop production, milk yield, and livelihood options in semi-arid regions of central India. Agronomy Journal, 114, 1769–1781. https://doi.org/10.1002/agj2.21062
10.1002/agj2.21062
CAS Web of Science® Google Scholar
Kumbhare, S. L., Sharma, K. N. S., & Patel, R. K. (1983). Standardisation of bovine units. Indian Journal of Animal Science, 53, 543–547.
Google Scholar
Leoni, F. M., Wasson, D. E., & Hristov, A. N. (2022). Feeding dairy cows for improved metabolism and health. Animal Frontiers, 12, 29–36. https://doi.org/10.1093/af/vfac059
10.1093/af/vfac059
PubMed Google Scholar
Maina, K. W., Ritho, C. N., Lukuyu, B. A., & Rao, E. J. O. (2020). Socio-economic determinants and impact of adopting climate-smart Brachiaria grass among dairy farmers in eastern and western regions of Kenya. Heliyon, 6, e04335. https://doi.org/10.1016/j.heliyon.2020.e04335
10.1016/j.heliyon.2020.e04335
CAS PubMed Web of Science® Google Scholar
Mwangi, M., & Kariuki, K. (2015). Factors determining adoption of new agricultural technology by smallholder farmers in developing countries. Journal of Economics and Sustainable Development, 6, 208–216.
Google Scholar
Pathak, P. S., Singh, J. P., Sharma, P., Singh, K. K., Dwivedi, P. N., &Singh, J. B. (2005). Bundelkhand region: Agricultural perspectives (Status, Constraints and Prospects). ICAR-Indian Grassland and Fodder Research Institute, Jhansi-284 003.
Google Scholar
Roy, A. K., Agrawal, R. K., Bhardwaj, N. R., Mishra, A. K., & Mahanta, S. K. (2019). Revisiting national forage demand and availability scenario. In A. K. Roy, R. K. Agrawal, & N. R. Bhardwaj (Eds.), Indian fodder scenario: Redefining state wise status (pp. 1–21). ICAR- AICRP on Forage Crops and Utilization.
Google Scholar
Saran, S., Singh, R. A., Singh, R., Rani, S. I., & Singh, K. K. (2000). Feed resources for rearing livestock in the Bundelkhand region of Uttar Pradesh. Indian Journal of Animal Science, 70, 526–529.
Google Scholar
Saxena, R., Choudhary, B. B., & Joshi, D. (2019). Enhancing farmers' incomes from livestock in eastern India with special emphasis on dairying: Major drivers and strategies. Indian Journal of Dairy Science, 72, 659–667. https://doi.org/10.33785/IJDS.2019.v72i06.012
10.33785/IJDS.2019.v72i06.012
Google Scholar
Sharma, P., Choudhary, B. B., Singh, P., Kumar, S., Gupta, G., & Dev, I. (2021). Can forage technologies transform Indian livestock sector? Evidences from smallholder dairy farmers in Bundelkhand region of central India. Agricultural Economics Research. Review, 34, 73–82. https://doi.org/10.5958/0974-0279.2021.00016.1
10.5958/0974-0279.2021.00016.1
Google Scholar
Sharma, P., Pandey, S., Meena, B. S., & Singh, N. P. (2007). Knowledge and adoption of livestock feeding practices in Jhansi district of Bundelkhand. Indian Journal of Dairy Sciences, 60, 63–67.
Google Scholar
Singh, P., Choudhary, B. B., Dwivedi, R. P., Arunachalam, A., Kumar, S., & Dev, I. (2023). Agroforestry improves food security and reduces income variability in semi-arid tropics of central India. Agroforestry Systems, 97, 509–518. https://doi.org/10.1007/s10457-023-00806-6
10.1007/s10457-023-00806-6
Google Scholar
Takahashi, K., & Barrett, C. B. (2013). The system of rice intensification and its impacts on household income and child schooling: Evidence from rural Indonesia. American Journal of Agricultural Economics, 96, 269–289. https://doi.org/10.1093/ajae/aat086
10.1093/ajae/aat086
Web of Science® Google Scholar
Wooldridge, J. M. (2010). Econometric analysis of cross section and panel data ( 2nd ed.). MIT press.
Google Scholar

Volume95, Issue1

January/December 2024

e70009

Modeling impact of improved forage cultivation on milk productivity and feed sufficiency in semiarid tropics of central India: A doubly robust analysis

Abstract

1 INTRODUCTION