1. Introduction: Agricultural systems are under increasing sustainability and environmental pressure to improve resource-use efficiency in response to rising energy demand, environmental degradation, and greenhouse gas emissions. Intensive use of fertilizers, fossil fuels, and irrigation water has increased production costs and reduced environmental sustainability. Improving resource-use efficiency is therefore essential for sustainable agricultural production. Data Envelopment Analysis (DEA), particularly the Constant Returns to Scale (CRS) and Variable Returns to Scale (VRS) models are an effective tool for evaluating technical efficiency and identifying input waste without reducing output. Integrating energy analysis with environmental indicators such as carbon efficiency and greenhouse gas emissions provides a comprehensive framework for sustainability assessment. Zucchini (Cucurbita pepo L.) is widely cultivated in Dezful County, Iran, under intensive, input-based systems that require substantial energy inputs. Excessive use of chemical fertilizers, diesel fuel, and electricity in these production systems contributes to low energy efficiency and high environmental impacts. This study evaluates energy use, efficiency indicators, GHG emissions, carbon performance, and sustainability indices in zucchini production, and determines optimal input-use patterns using DEA.
2. Materials and Methods: The study was conducted on 30 zucchini farms in Dezful County, Khuzestan Province, Iran. Data on inputs, including human labor, machinery, diesel fuel, fertilizers, pesticides, irrigation water, electricity, and seed, were collected through field surveys. All inputs and outputs were converted into energy units (MJ ha⁻¹) using standard energy equivalents. Energy indicators, including total input and output energy, energy ratio, net energy, and specific energy, were calculated. DEA input-oriented models under CRS and VRS assumptions were applied to evaluate technical efficiency and estimate potential input savings. Environmental impacts were assessed by calculating emissions of CO₂, CH₄, and N₂O from agricultural inputs. Global Warming Potential (GWP) was expressed as CO₂ equivalents. Carbon input, carbon output, net carbon gain, carbon efficiency, and sustainability index were also computed. Results from conventional and optimized scenarios were compared to evaluate improvements in energy and environmental performance.
3. Results and Discussion: Total energy input and output in conventional production were 41,414.28 and 48,042.74 MJ ha⁻¹, respectively. Electricity (31.62%), nitrogen fertilizer (28.91%), diesel fuel (16.72%), and human labor (6.48%) were the main contributors to energy consumption, indicating that irrigation and fertilizer management are key factors in energy use. DEA results showed significant inefficiencies in input use. Energy consumption decreased from 41,414.28 MJ ha⁻¹ in the conventional system to 33,973.29 MJ ha⁻¹ under the optimized scenario. The highest energy savings were related to nitrogen fertilizer (39.91%), followed by pesticides (31.07%), human labor (24.21%), and diesel fuel (21.21%). These results highlight the potential for improving farm management practices and reducing unnecessary input use. Energy performance indicators improved under the optimized system. Specific energy decreased, indicating lower energy use per unit of output. Energy ratio and net energy increased under the optimized scenario, indicating more efficient utilization of energy inputs and greater energy productivity per unit of resource consumed.
Environmental analysis revealed a reduction in total GWP from about 5,185.81 kg CO₂eq ha⁻¹ in the conventional system to 4,249.94 kg CO₂eq ha⁻¹ in the optimized system. Electricity was the largest contributor to emissions (>70%), followed by diesel fuel and nitrogen fertilizer. Diesel-related emissions decreased significantly due to improved machinery management, while reduced nitrogen application lowered soil-related N₂O emissions. Despite improvements, irrigation-related electricity use remained the main environmental concern. DEA-based efficiency analysis indicated that 16 farms were fully efficient under CRS, while 22 farms were efficient under VRS. Average efficiency scores were 96.61% (CRS) and 99.38% (VRS). Inefficient farms could reduce input use by about 7.25% without affecting yield. The VRS model identified higher potential savings due to its flexibility in accounting for scale differences. Carbon analysis showed that carbon output remained constant at 27,024.04 kg C ha⁻¹, indicating that production levels were maintained. Carbon input decreased from 1,400.17 to 1,147.48 kg C ha⁻¹, increasing net carbon gain. Carbon efficiency improved from 19.31 to 23.56, while the sustainability index increased from 18.35 to 22.55. These improvements confirm enhanced environmental performance under optimized management.
4. Conclusion: This study demonstrated that DEA-based optimization can substantially improve the energy and environmental performance of zucchini production systems without reducing yield. Electricity, diesel fuel, and nitrogen fertilizer were identified as the major contributors to energy use and emissions. Significant reductions in energy consumption and greenhouse gas emissions were achieved through more efficient use of production inputs, while improvements in energy indicators, carbon efficiency, and sustainability indices confirmed the effectiveness of the optimized management scenario. It was observed that improving irrigation efficiency, optimizing fertilizer application, reducing fossil fuel use, and adopting renewable energy systems can further enhance the sustainability of zucchini production. Overall, integrating energy analysis with DEA provides a powerful approach for improving resource efficiency and supporting sustainable agricultural development.