The scale of the data collection alone sets this work apart. Working with National Agricultural Research and Extension Services partners, the team surveyed farmers in Burkina Faso, Burundi, the Democratic Republic of the Congo, Kenya, Madagascar, Nigeria, Rwanda, Sierra Leone, Tanzania, and Uganda after the 2020/2021 harvest season. The SRP survey tool covered all 46 requirements of the SRP Standard and its 12 performance indicators, a standardized, farm-level framework spanning economic, environmental, and social dimensions of sustainability. Of the households surveyed, 1,039 were managed by women and 2,042 by men, selected through multistage random sampling across three distinct production systems: irrigated lowland, rainfed lowland, and rainfed upland environments. The researchers focused on five performance indicators for agronomic gain: grain yield, net profit, labor productivity, and the partial factor productivity of nitrogen (PFPN) and phosphorus (PFPP).

The resource disparities documented in the survey are stark. Female-managed farms operated plots that were 36 to 41 percent smaller than those run by men, and they used roughly half the agricultural equipment. The share of female-managed farms varied dramatically by country, from just 8 percent in Nigeria’s irrigated lowlands to 45 percent in Sierra Leone and 44 percent in Madagascar. Fertilizer use told a similar story of inequity. Average nitrogen application rates reached 63 kilograms per hectare in irrigated lowlands but dropped to 23 and 18 kilograms per hectare in rainfed lowland and upland systems respectively, and women farmers generally applied less fertilizer than men, an average of 35 versus 54 kilograms of nitrogen per hectare when comparing female-managed to male-managed operations in some contexts. In Uganda, Sierra Leone, and Madagascar, surveyed farmers applied no nitrogen fertilizer at all.

Yields reflected these unequal inputs. In rainfed systems, 82 percent of female-managed farms and 72 percent of male-managed farms produced 3 tons per hectare or less, and only 7 percent of all farms exceeded 5 tons per hectare. Average yields ranged from a low of 0.6 tons per hectare in rainfed lowlands of Sierra Leone to 7.2 tons per hectare in Tanzania’s irrigated systems. Across all sites, female-managed farms averaged 3.1 tons per hectare compared with 4.1 tons for male-managed farms, a statistically significant difference. Male-managed farms consistently outperformed in irrigated lowlands (4.8 versus 4.0 tons per hectare), rainfed lowlands (2.5 versus 2.1), and rainfed uplands (2.4 versus 2.0).

Perhaps the most striking results concern labor productivity and profitability gaps. The researchers calculated gaps as the percentage difference between the top 10th percentile of performance within each country and production system and the overall mean. Yield gaps averaged 38 percent in irrigated lowlands, 48 percent in rainfed lowlands, and 53 percent in rainfed uplands. Net profit gaps exceeded 50 percent in all three systems, averaging 56, 57, and 67 percent respectively. Labor productivity gaps were the widest of all, averaging 72 percent in irrigated lowlands. When disaggregated by gender, female-managed farms showed significantly higher labor productivity gaps than male-managed farms in every production system, with gaps ranging from 78 to 99 percent in individual countries, compared with 3 to 68 percent for male-managed farms. In Nigeria, for instance, the female labor productivity gap reached 98 percent in rainfed lowlands and 99 percent in rainfed uplands.

Labor allocation itself followed gendered patterns. On female-managed farms, female workers contributed an average of 35 person-days per hectare per season, compared with 19 person-days on male-managed farms, while male labor input was greater on male-managed farms at 171 versus 160 person-days per hectare. Notably, when labor productivity was disaggregated by the gender of both the worker and the manager, female labor was more productive on female-managed farms than on male-managed farms, and the reverse held for male labor. The authors suggest this gender-concordant efficiency may reflect differences in labor organization, supervision, and task allocation, with managers optimizing the performance of work groups with which they share roles, experience, or communication dynamics.

To identify what drives performance, the team turned to machine learning, fitting random forest models with 500 trees and then quantifying the influence of the top predictors through multiple regression. Nitrogen input, total labor input, plot size, phosphorus input, and the number of irrigations emerged as the leading drivers of yield for both female- and male-managed farms, and the models showed their strongest predictive power for yield, with R-squared values between 0.63 and 0.67. For net profit, the top five predictors on female-managed farms were the use of certified seeds, adherence to a crop calendar, labor input, plot size, and the number of equipment items, while on male-managed farms the list included nitrogen input and training in sustainable practices. However, the models performed poorly for profit (R-squared of 0.13 to 0.16), suggesting that profitability is shaped by factors beyond the farm-level variables measured, such as market conditions and price structures. For labor productivity, plot size and labor input were the two primary drivers for both groups, with training in sustainable practices and record keeping playing significant roles, particularly on female-managed farms.

Nutrient use efficiency added nuance to the picture. The partial factor productivity of nitrogen ranged from 25 to 127 kilograms of grain per kilogram of elemental nitrogen across countries, while phosphorus efficiency ranged from 13 to 642 kilograms of grain per kilogram of elemental phosphorus. Only PFPN differed significantly between female- and male-managed farms in the lowland systems, with women’s farms averaging 65 versus men’s 70 kilograms of grain per kilogram of nitrogen overall. Interestingly, in rainfed systems, more female-managed farms fell above the desirable nitrogen efficiency threshold of 80 kilograms of grain per kilogram of nitrogen, suggesting that although women apply less fertilizer, they may use what little they have relatively efficiently. Multinomial logistic regression showed that higher nitrogen input reduced the likelihood of both abnormally low and abnormally high nitrogen efficiency on female-managed farms, while phosphorus input and equipment access stabilized phosphorus efficiency, pointing to the importance of balanced, site-specific nutrient management rather than input intensification alone.

The implications, the authors argue, extend well beyond agriculture statistics. Closing gender gaps in agricultural productivity could add roughly $1 trillion to global gross domestic product and lift 45 million people out of hunger, according to estimates cited in the study. In sub-Saharan Africa, where domestic rice production currently meets only 60 percent of consumption and demand is expected to reach approximately 150 million tons by 2050, leaving half the farming population systematically underperforming represents an enormous lost opportunity. The study’s findings that female-managed farms had larger gaps across yield, profit, and labor productivity in nearly every country and system indicate that the barriers are structural rather than a matter of farming ability. Customary land tenure systems frequently assign women smaller and less fertile plots, and the more productive lowland environments, where water control boosts yields, are often under men’s control.

The researchers are careful to note the limitations of their work. The analysis rests on a single cross-sectional survey season, precluding causal inference, and net profit calculations excluded the cost of family labor, which may overstate profitability, particularly on female-managed farms where unpaid household labor contributes substantially. Data on education, market access, credit availability, and land tenure security were incomplete or absent, and self-reported inputs carry the risk of recall bias. Nevertheless, the breadth of the survey, ten countries, three production systems, and more than 3,000 households, provides an unusually robust comparative foundation.

The policy prescriptions that emerge are equally clear-eyed. The authors conclude that improving women’s access to inputs alone will not close gender gaps; instead, an integrated approach combining fertilizer access with nutrient management training, responsive extension services, labor-saving technologies, inclusive financing, and improved access to land and equipment is needed. They also recommend embedding gender-disaggregated indicators into monitoring frameworks such as the Sustainable Rice Platform itself, so that interventions can be evaluated not only for whether they boost sustainability overall, but for whether they narrow the persistent gaps between women and men who grow one of the world’s most important staple crops. As rice demand surges across Africa, the study makes a compelling case that the continent’s food security may depend on how well its agricultural systems unleash the potential of its women farmers.

Traditional methods of sustainability assessment in agriculture have often been fragmented or overly narrow, focusing either exclusively on environmental indicators—such as soil health, water quality, and biodiversity—or solely on economic viability and social factors like farmer income and community resilience. The ASI distinguishes itself through its integrative framework, harmoniously blending ecological parameters with socioeconomic metrics, thereby capturing the multidimensional realities of agroecosystems. This comprehensive approach aligns closely with modern sustainability science’s call for multidisciplinarity and holistic evaluation.

At its core, the ASI synthesizes a diverse array of data points collected from field measurements, remote sensing technologies, and social surveys. Environmental dimensions incorporated in the index include soil fertility, greenhouse gas emissions, water consumption, and biodiversity indices focusing on pollinator presence and pest regulation. Meanwhile, socioeconomic dimensions assess farmer livelihoods, equity in resource access, community participation in governance, and market resilience. By marrying these datasets, the ASI produces an accessible yet nuanced single score representing the overall sustainability status of a given agroecosystem.

The development of the ASI was driven by a crucial need: to produce a metric not only scientifically robust and translatable across diverse agricultural contexts but also practical for stakeholders ranging from local farmers to international agencies. Importantly, the authors designed the tool to be adaptable, allowing incorporation of region-specific parameters while maintaining a unified core framework to facilitate standardized comparison. This paves the way for novel insights into how different agrarian models—from smallholder farms in sub-Saharan Africa to industrial row cropping in North America—perform on sustainability.

The methodology underlying the ASI involved comprehensive field campaigns across multiple continents, encompassing varied crop systems and management practices. The researchers employed advanced statistical modeling and machine learning algorithms to validate indicator selection and weighting, enhancing the index’s predictive power and reliability. Rigorous cross-validation ensured that the ASI accurately reflects real-world conditions and outcomes related to sustainability goals outlined by the UN Sustainable Development Goals (SDGs), especially those targeting zero hunger, clean water, climate action, and responsible consumption.

One particularly innovative feature of the ASI includes its dynamic temporal component. Unlike static sustainability assessments, the index can capture changes over time, thus enabling the monitoring of progress or decline in agroecosystem health and social well-being. Temporal analysis is critical for evaluating the impact of interventions, policy changes, and emerging environmental threats such as drought or pest outbreaks. This time-sensitive capability transforms the ASI into a proactive tool, guiding adaptive management strategies and investment priorities.

From an environmental science perspective, the ASI’s emphasis on biodiversity and soil health is especially noteworthy. Soil organic carbon levels and microbial activity, key indicators of soil vitality, are integrated alongside landscape-level biodiversity metrics encompassing native flora and fauna diversity. By quantifying these elements, the ASI addresses the core ecological functions that underpin productive and resilient farming systems. This approach reflects a paradigm shift recognizing that agroecosystems are not mere food-production units but complex socioecological entities requiring balanced stewardship.

Simultaneously, the socioeconomic component delves into the livelihoods and rights of farming communities, a historically underrepresented domain in sustainability assessments. The index evaluates factors such as income stability, access to technology and credit, gender equity, and the inclusiveness of decision-making processes. This illuminates how economic and social equity interconnect with ecological outcomes, reinforcing that sustainability extends beyond environmental metrics to encompass justice and human dignity within agricultural livelihoods.

In practical applications, preliminary deployments of the ASI have already begun revealing striking patterns. In one case study focusing on Mediterranean agroecosystems, the tool helped identify critical trade-offs where intensification boosted short-term yields but compromised long-term soil health and social cohesion. Such insights spotlight the urgency of recalibrating agricultural practices to embrace regenerative principles. The ASI also aids certification bodies and sustainability labeling programs by supplying scientifically rigorous benchmarks to support transparency and consumer awareness.

Importantly, the ASI holds profound implications for climate resilience. By examining greenhouse gas emissions alongside adaptive capacity indicators—such as diversification of income sources and community networks—the index becomes a litmus test for agroecosystem vulnerability in the climate crisis. Policymakers can harness this data to channel resources toward regions and practices that not only mitigate carbon footprints but also bolster smallholder resilience against extreme weather and market volatility.

The study also delves into the computational architecture facilitating ASI use. Developed with an open-source platform, the index is accessible to a broad array of users, including researchers, NGOs, and local governments, promoting widespread adoption and collaborative improvement. By embedding machine learning capabilities, the tool continuously evolves as new data accrues, ensuring sustained relevance amid rapidly shifting agricultural and environmental conditions.

Critically, the authors highlight that the ASI should not be viewed as a static verdict but rather as a dynamic guide to sustainability trajectories. Engaging with farmers and communities in interpreting ASI results is fundamental to the tool’s success, fostering participatory approaches that empower stakeholders to co-create sustainable futures. This engagement also mitigates risks of techno-centric reductionism, ensuring that the index remains grounded in local realities and knowledge systems.

The introduction of the Agroecosystem Sustainability Index aligns with a broader scientific momentum to redefine sustainability beyond rhetoric and fragmented measures. As agriculture stands at the nexus of food security, environmental degradation, and socioeconomic inequality, the ASI offers a pathway toward more nuanced and actionable understanding. Its capacity to reflect intertwined ecological and social dimensions promises to underpin transformative policies and practices essential for meeting global sustainability challenges.

In conclusion, the publication of this innovative ASI framework arrives at a critical juncture, providing a much-needed compass in the quest for sustainable agriculture. By delivering a robust, flexible, and comprehensive metric, the work of Mühlematter, Maund, and Nina equips the global community with powerful insights needed to balance productivity with planetary and societal health. The ASI exemplifies how multidisciplinary collaboration and methodological innovation can usher in a new era where agroecosystem management harmonizes human prosperity with ecological stewardship.

Subject of Research:

Article Title:

Article References:
Mühlematter, D.J., Maund, S.J. & Nina, M. Agroecosystem sustainability index ASI for measuring environmental and socioeconomic sustainability. npj Sustain. Agric. 3, 51 (2025). https://doi.org/10.1038/s44264-025-00095-9

Image Credits: AI Generated