The research team, led by Debasmita Baruah of the Department of Agricultural Economics at Assam Agricultural University in Jorhat, together with colleagues from institutions including the ICAR National Institute of Agricultural Economics and Policy Research in New Delhi, analysed climatic data spanning 1981 to 2024. Rather than relying on a single summary statistic, the researchers examined four separate parameters that matter directly to agriculture: rainfall, maximum temperature, minimum temperature, and relative humidity. Each was tested for long-term trend using the Mann-Kendall test, a non-parametric method widely used in hydrology and climate science because it makes no assumptions about the underlying distribution of the data and is robust to outliers, and the magnitude of any trend was quantified with Sen’s slope estimator, which computes the median of all pairwise slopes in the series.

The statistical verdict was clear on two fronts. Minimum temperature rose at a rate of 0.034 degrees Celsius per year, a trend that was statistically significant and, compounded over the 43-year record, amounts to a shift of well over a degree in nighttime conditions. Relative humidity increased by 0.18 percentage points per year, also a significant trend. Rainfall, by contrast, showed a positive direction but failed to reach statistical significance, meaning the data cannot rule out that the apparent increase is noise. This asymmetry matters agronomically. Warmer nights accelerate crop respiration, erode the accumulation of dry matter in grains, and raise the vapour pressure deficit stress on plants, while persistently higher humidity favours fungal and bacterial pathogens that attack rice, vegetables, and pulses. A climate that is not dramatically hotter by day can still be substantially more hostile to farming.

Assam’s agriculture is dominated by small and marginal holdings, and the state’s position in the Brahmaputra valley exposes it to floods, erosion, and increasingly erratic monsoon behaviour. To understand how farmers are responding, the team turned to primary data collected from 300 farmers using a multistage stratified random sampling technique, a design that first divides the population into meaningful strata, in this case across three agro-climatic zones, and then samples randomly within them at successive stages. The survey focused on the Central Brahmaputra Valley Zone, and the authors acknowledge the farmers of that zone for their cooperation. Ethical approval for the survey was obtained from the Institutional Ethics Committee of Assam Agricultural University, and informed consent was taken from all participants.

From the survey responses, the researchers constructed an adaptation index, a composite score that measures the extent to which each farmer has adopted recognised climate-change adaptation practices. The results were striking in their hierarchy. Crop diversification emerged as the most widely adopted strategy, with an index score of 0.99, meaning it was essentially universal among the sampled farmers. Spreading production across multiple crops is a classic risk-spreading manoeuvre: if floods or pests wipe out one crop, others may survive, smoothing household income and food supply in the face of climatic shocks. Off-season vegetable cultivation followed with a score of 0.87, a strategy that both exploits market windows when prices are favourable and shifts production away from the periods of highest climatic risk. Crop-livestock integration ranked third at 0.71, providing farmers with a buffer of animal products and manure that decouples part of household income from the fate of the fields.

But knowing what farmers do is only half the question. The more policy-relevant puzzle is who adapts and why. To answer it, the team employed binary logistic regression, a statistical model suited to a binary outcome, in this case whether or not a farmer has adapted to climate change. The model estimates the log-odds of adaptation as a function of socio-economic explanatory variables, allowing each factor’s effect to be isolated while holding the others constant. The results pointed unambiguously to two drivers. Farm size was highly significant, with a p-value below 0.001, and extension contact, the number of times a farmer interacts with agricultural advisory services per year, was also significant at p equal to 0.007.

The effect sizes translate into concrete, actionable numbers. A one-hectare increase in farm size was associated with an approximately 16 percentage-point increase in the probability that a farmer adapts. One additional extension contact per year raised the predicted probability of adaptation by roughly 0.97 percentage points, all else being equal. These figures carry a double message. First, larger landholders have both the resources to invest in new practices and the cushion to absorb the risk of trying them, so adaptation capacity is stratified by wealth in exactly the way development economists would predict. Second, and more optimistically, extension contact matters independently of land size, which means that the flow of information from agricultural scientists and field officers to farmers is a genuine lever that policy can pull, and one that does not require redistributing land.

There is a sobering arithmetic hidden in that 0.97 percentage-point figure. If a single extra advisory contact per year buys less than one percentage point of adaptation probability, then moving the needle for a large share of Assam’s farming population will require a sustained, multiplicative expansion of extension services, training programmes, and climate-smart technology delivery, not a one-off campaign. The authors argue that improving farmers’ access to extension services, training, and climate-smart technologies is the key local-level priority. At the regional level, they note, the findings can help policymakers in areas with similar agro-climatic conditions design strategies to strengthen agricultural resilience and food security, a point of considerable relevance across the eastern Himalayan foothills and the densely farmed river deltas of South Asia, where millions of smallholders face comparable combinations of warming nights, humid air, and erratic monsoons.

What makes the study methodologically interesting is the way it stitches together two kinds of evidence that rarely meet in a single paper. On one side sits a 43-year instrumental record interrogated with trend tests that are standard precisely because they are conservative; on the other, a structured household survey converted into a quantified adaptation index and pushed through a regression model with interpretable marginal effects. The connection between the two is the crux. The climate data establish that the hazard is real and directional, while the survey data establish that the response is uneven and socially patterned. A significant rise in minimum temperature and humidity is not, by itself, a crisis; it becomes one when the farmers least equipped to respond, those with the smallest plots and the thinnest connections to advisory networks, are the ones most exposed.

The broader lesson extends well beyond Assam. Around the world, adaptation to climate change is often discussed as if it were a purely technical problem, a matter of breeding hardier varieties or forecasting floods better. This study is a reminder that adaptation is also a distributional problem. The practices that buffer farmers against climatic risk, diversifying crops, shifting planting calendars, integrating livestock, are already known and already adopted by many. What separates adapters from non-adapters is not knowledge of the practices but the land and the information channels that make adoption feasible. As nighttime temperatures continue their statistically significant climb and humidity keeps rising across the Brahmaputra valley, the study suggests that the cheapest and fastest intervention available to governments may not be a new technology at all, but a more frequent knock on the farmer’s door from someone carrying useful, locally relevant advice. In a state where agriculture remains the backbone of rural livelihoods, that knock could determine who weathers the coming decades and who does not.

Subject of Research: Climate trends and farmer adaptation strategies across agro-climatic zones of Assam, India

Article Title: Climate change and agricultural resilience in Assam: unravelling climate trends, farmer adaptation, and socioeconomic drivers across agro-climatic zones

Article References: Baruah, D., Kishore, P., Rede, G. D., Barman, U., Saikia, H., Majumder, S. H., & Deka, N. (2026). Climate change and agricultural resilience in Assam: unravelling climate trends, farmer adaptation, and socioeconomic drivers across agro-climatic zones. Discover Sustainability. https://doi.org/10.1007/s43621-026-05005-8

Image Credits: AI Generated

DOI: 10.1007/s43621-026-05005-8

Keywords: climate change, agricultural adaptation, Assam, agro-climatic zones, Mann-Kendall test, Sen's slope, crop diversification, extension services, smallholder farmers, logistic regression, agricultural resilience, food security