Introduction
Global population growth and shrinking natural resources have intensified pressure on agricultural systems, positioning sustainable waste management as a national priority, particularly in arid and semi-arid countries like Iran. Agricultural activities generate substantial biomass residues, yet inefficient handling (e.g., open burning, uncontrolled dumping) contributes significantly to greenhouse gas (GHG) emissions (CO₂, CH₄), energy overuse, and economic losses. Globally, one-third of food production is lost or wasted (FAO, 2011, 2013), with profound environmental and food security implications. In this context, the Circular Waste-Based Bioeconomy (CWBE) framework offers a strategic pathway: integrating circular economy principles into agricultural systems to transform waste into compost, biogas, animal feed, and bio-based materials—thereby enhancing resource efficiency and reducing environmental footprints. However, existing efficiency assessments in agricultural waste management in Iran predominantly ignore undesirable outputs, yielding an incomplete picture of sustainability. This study addresses that gap by evaluating the Eco-efficiency of agricultural waste recycling across Iran’s 31 provinces over 2014–2019 (1393–1398), using a Dynamic Network Data Envelopment Analysis (DN-DEA) model that jointly accounts for desirable and undesirable outputs.

Materials and Methods
We construct a dynamic network model inspired by Lu et al. (2022), conceptualizing waste recycling as a two‑stage system: (1) waste collection and preprocessing, and (2) valorization and environmental mitigation. The DN‑DEA model allows efficiency scores to reflect carry‑over effects—such as investment in fixed assets—across time periods, making it suitable for assessing policy persistence and technological lock‑in.
Inputs include: (i) monetary cost of managing agricultural waste (IRR, adjusted for inflation), and (ii) energy consumption (GJ) in collection, transport, and processing. Desirable output is the quantity of recycled biomass (tonnes), while undesirable outputs encompass: (a) CO₂ emissions (tonnes), (b) CH₄ emissions (tonnes), (c) total energy consumption (as above, dual‑counted to penalize inefficiency), and (d) waste management cost (as a proxy for incomplete valorization). Data were compiled from the Statistical Center of Iran, Agricultural Jihad provincial reports, and the Central Bank’s economic databases. All monetary values were deflated using the consumer price index (base = 1390), and emission factors were standardized using IPCC Tier 1 guidelines adapted to Iran’s agronomic conditions. The model was solved using GAMS 36.2.0 (with CONOPT solver), and statistical validation—including Pearson correlation matrices and significance tests—was performed in SPSS 28.
Results and Discussion
Results reveal striking regional heterogeneity. Five provinces—Tehran, Bushehr, Ilam, Khuzestan, and South Khorasan—consistently achieved full efficiency (score = 1.00) across all six years, functioning as benchmarks. In contrast, Golestan (avg. θ = 0.34), Kurdistan (0.35), Sistan & Baluchestan (0.36), East Azerbaijan (0.41), and Kermanshah (0.41) exhibited chronic underperformance. Nationally, average efficiency declined from 0.72 in 2014 to 0.61 in 2019, signaling systemic deterioration—potentially linked to subsidy distortions, energy price volatility, or drought-induced feedstock scarcity.
Crucially, Pearson correlation analysis confirmed strong, positive, and statistically significant (p < 0.01) relationships between recycling efficiency and the management of undesirable outputs. This implies eco-efficiency synergy: provinces excelling in one dimension (e.g., lowering disposal costs) tend to outperform across all environmental and economic metrics. Moreover, undesirable outputs are highly inter correlated (e.g., energy–CO₂: r = 1.000; CO₂–CH₄: r = 0.999), suggesting that interventions targeting energy efficiency will likely reduce multiple emissions simultaneously. Spatial mapping (Figure 2) further confirms a pronounced east–west efficiency gradient, with high-efficiency clusters in central/southern/eastern provinces and low-efficiency zones concentrated in the northwest and southwest—mirroring disparities in infrastructure, technology access, and institutional capacity.

Conclusion
This study advances the literature by: (i) applying DN-DEA to Iranian agricultural waste for the first time; (ii) integrating four undesirable outputs simultaneously; and (iii) linking efficiency diagnostics to concrete, region-specific policy levers. Findings underscore that a one-size-fits-all approach is inadequate: instead, precision sustainability governance, where efficiency metrics inform spatially differentiated interventions, is essential. We recommend:
Region-targeted interventions for low-performing provinces (e.g., Golestan, Kurdistan, Sistan & Baluchestan).
Knowledge diffusion mechanisms, such as peer-learning networks, to transfer best practices from benchmark provinces
Efficiency-based budgeting, where provincial subsidies or green credit access are tied to annual DN-DEA scores.
Ultimately, improving eco-efficiency in agricultural waste recycling is not merely a technical challenge but a governance imperative, one that can simultaneously bolster rural livelihoods, enhance resource security, and advance Iran’s commitment to low-carbon development.