Sustainable grazing management supports climate change mitigation in a mediterranean pasture

Abstract

Mediterranean dryland sheep farms face interconnected challenges of high greenhouse gas emission intensity and progressive land degradation driven by continuous, unmanaged grazing. Rotational grazing (RG) has been proposed as an integrated response, but field evidence documenting the climate outcomes of real farm-level grazing management transitions remains scarce. This study quantifies the carbon footprint (CF) of a commercial Montado sheep farm in Mediterranean drylands (Alentejo, Portugal) before and after a management transition implemented in 2018, using a cradle-to-farm-gate life cycle assessment and 100-year global warming potentials from the Sixth Assessment Report of the Intergovernmental Panel on Climate Change. Prior to 2018, the farm operated under continuous grazing with documented land degradation and soil organic carbon below 2%. In 2018, the farm converted to RG through paddock restructuring, reduced stocking pressure, and a progressive reduction of external feed and fertilizer inputs. Total farm emissions were 34% lower in the post-transition period (1,165.1 t CO₂-eq yr^-1) than pre-transition (1,756.6 t CO₂-eq yr^-1), driven by reduced feed, fertilizer, and fuel inputs. Farm operations showed the largest reduction (59%), while enteric fermentation remained the dominant source in both periods. The gross CF of lamb meat decreased from 16.2 to 12.7 kg CO₂-eq kg^-1 live weight (LW) following the transition. Post-transition soil carbon (C) accrual was estimated at 2.47 to 2.9 t C ha^-1 yr^-1 across 200 ha of pasture, consistent with early-phase recovery from a severely degraded, low-carbon baseline. When incorporated into net C accounting, the farm shifted to a C sink, giving a net carbon footprint of −7.1 kg CO₂-eq kg^-1 LW after accounting for soil carbon accrual. These findings demonstrate that a whole-farm grazing management transition can simultaneously reduce emission intensity, generate measurable soil C accrual, and deliver ecosystem services, with direct relevance to C farming policy and climate strategy for Mediterranean dryland livestock systems.

1 Introduction

Land degradation and livestock-related greenhouse gas (GHG) emissions are two key sustainability issues associated with pasture management in the Mediterranean region. Decades of continuous, conventional grazing across the region, combined with intensifying climate change towards drier conditions, as well as increased fire risk, have progressively depleted soil organic carbon (SOC), destabilized pasture vegetation, decreased the carbon (C) mitigation potential of pastures, and increased desertification risk (Kairis et al., 2015; Kosmas et al., 2015; Ferreira et al., 2022). In addition, the increase in herd size, observed in Mediterranean extensive pastoral systems in the past decades (Delattre et al., 2020), has increased the risk of pasture degradation and the dependence on external feed inputs (Delattre et al., 2020; Grilli et al., 2021; Sabater et al., 2022; Roxo et al., 2023). Agricultural intensification and soil degradation contribute to increasing the net GHG emissions balance of pastures (GHG per ha). In 2023, livestock emissions (4.3 Gt CO₂-eq) were the largest single component of the global agrifood system (16.5 Gt CO₂-eq), followed by deforestation (2.8 Gt CO₂-eq), and packaging, transport and retail (1.4 Gt CO₂-eq) (FAO, 2025).

The European Union (EU) Carbon Removals and Carbon Farming Certification (CRCF) Regulation (EU 2024/3012), adopted in November 2024, establishes a framework for certifying and rewarding agricultural C removals alongside GHG emission reductions, paving the way for stronger adoption of more sustainable pasture management strategies by farmers and meat producers (European Parliament and Council, 2024). The CRCF also points out that these decarbonization targets need to be reached whilst improving other key nature benefits like soil health and biodiversity. At the same time the New EU Agricultural Vision 2040 is paving the way for more sustainable agriculture introducing agroecological solutions into land management, with less external inputs and xenobiotics, higher circularity, attention to biodiversity and soil health.

Extensive Mediterranean pastures pose in this respect a very interesting playground as the “land-animal” integration system allows to explore sustainable management solutions that support self-sufficiency and reduce dependency on external inputs, leading to win-win outputs. These benefits include increased soil C accrual through reduced erosion and enhanced C sequestration, improved soil ecosystem services, and reduced land degradation and desertification risk. The introduction of sustainable management strategies is also key for climate adaptation of these productive systems that will require healthy ecosystem conditions to continue to support land productivity under increasingly harsh environmental conditions.

Rotational grazing (RG) is a land management strategy with documented positive effects on pasture quality and soil health. In contrast, continuous or poorly managed grazing may contribute to pasture degradation, soil compaction, and to the loss of soil organic matter (Oggioni et al., 2020; Centeri, 2022; Zhang et al., 2022). The main principles of RG include subdividing grazing areas into paddocks and moving animals among them according to a planned schedule. This approach regulates grazing pressure, allows vegetation to recover between grazing events, and promotes SOC accumulation through increased root biomass inputs and reduced soil disturbance (Augustine et al., 2020; de Otálora et al., 2021; Teutscherová et al., 2021). Available syntheses confirm that improved grassland management can be particularly beneficial for C accrual on degraded soils with low initial C content (Conant et al., 2017; Bai and Cotrufo, 2022; Dondini et al., 2023) like those most frequently found in Mediterranean pastures (Grilli et al., 2021). Beyond C, well-managed rotational systems deliver a range of ecosystem services, including improved water infiltration, enhanced soil microbial activity, pasture productivity recovery, and biodiversity support. These outcomes are especially relevant in dryland agroecosystems under desertification risk (Horrillo et al., 2024; Vagnoni et al., 2024; Ishaq et al., 2025a). The Montado agro-silvo-pastoral system of southern Portugal, a mosaic of open-canopy cork oak woodland and grazed herbaceous understory (Capelo and Vila-Viçosa, 2021), is a clear example of this context: a multifunctional landscape with high ecological value yet highly vulnerable to degradation under continuous grazing (CG) pressure (Gaspar et al., 2008; Roxo et al., 2023).

This study examines a commercial sheep farm managed extensively for meat production in the pasturelands of Portugal. The system was managed under CG until 2018. The land was characterized by significantly low levels of organic matter and nutrients (Grilli et al., 2021) showing clear signs of land degradation and limited grass productivity, which obliged the landowner to rely on external feed at least during part of the year. In 2018, the farm, in collaboration with the LIFE project Desert-Adapt (LIFE16 CCA/IT/000011), introduced sustainable land management solutions, including RG, reduced stocking pressure, and progressively shifted away from external feed and fertilizer inputs (Life desert-adapt, 2026).

The goal of this study was to evaluate the climate mitigation potential of the introduced sustainable land management by comparing the total GHG balance of the productive system before and after the land management transition. We hypothesized that, in line with C farming principles, the new management would result in lower net farm-level CO₂-eq emissions and consequently reduce the carbon footprint (CF) of sheep meat, the farm’s main marketable product. Evidence of multiple environmental benefits associated with sustainable farming can support farmers’ adoption of sustainable practices and contribute to the EU agricultural transition towards its 2040 goals.

2 Materials and methods

2.1 Study area and management transition

The sheep farm was located in the Alentejo region, Portugal (Mértola Municipality, Beja District; 37° 43′51″N, 7° 43′38″W to 37° 48′41″N, 7° 39′49″W). The climate is warm temperate with hot dry summers (Köppen classification), mean annual precipitation of 456 mm, mean annual temperature of 16.2 °C, and an aridity index of 0.34. Soils are acidic (pH ∼6.0), developed on Schist and Quartzite substrates, with SOC generally below 2% at the time of the transition (Grilli et al., 2021). The dominant land use is Montado, an agro-silvo-pastoral mosaic of open-canopy evergreen oak woodland (Quercus suber L. and Q. coccifera L.) with a herbaceous understory maintained through extensive sheep grazing (Capelo and Vila-Viçosa, 2021).

Prior to 2018, the farm operated under CG across approximately 200 ha of pasture, supplemented by conventional cultivation of an additional 120 ha for feed production. In 2018 the farm transitioned to RG, which involved subdividing the main pasture into smaller paddocks using electric fencing (Ishaq et al., 2025a), reducing stocking pressure, and progressively integrating the 120 ha of cultivated land into the rotational scheme as forage productivity recovered, thereby reducing dependence on purchased feed and synthetic fertilizers during the post-transition period. The effect of the management transition on soil quality was assessed in 2018 and again in 2022, 4 years later, by measuring soil C and N contents and related physicochemical properties.

In order to assess the climate mitigation potential of the farm, as well as the C footprint of its products, we applied a life cycle assessment that accounted for all GHG emissions associated with land management and the different components of the production system, including changes resulting from reduced external inputs and lower N₂O emissions due to decreased N fertilizer use. Additionally, soil C accrual was included as a negative term in the overall farm-level CO_2-eq balance, following C farming accounting principles that consider both C sequestration and CO₂ and non-CO₂ GHG emissions, particularly CH₄ and N₂O. Enteric CH₄ emissions were also estimated, but as this study did not involve direct measurements on animals, these emissions were calculated using standard IPCC methods and input parameters. Animal performance parameters were held constant between the two management periods, whereas flock size and forage digestibility differed.

2.2 Life cycle assessment of farm management and products

We applied a life cycle assessment (LCA) (ISO, 2006a; ISO, 2006b) and CF quantification in accordance with ISO 14067 (ISO, 2018) and FAO Livestock Environmental Assessment and Performance (LEAP) guidelines for small ruminants (FAO, 2016). Analyses were performed in SimaPro v9.5 using the Ecoinvent v3.10 database. The system boundary was cradle-to-farm-gate, encompassing upstream processes for feed production, fertilizer and fuel manufacturing, and transport, as well as on-farm emissions from enteric fermentation, manure management, and field operations. Capital goods were included with impacts annualized over service life. The system boundary and the main organizational differences between the farming system before and after the land management introduction are illustrated in Figure 1.

FIGURE 1

Feed inputs were grouped into four categories: grazed pasture, on-farm roughage (hay and silage), grains (oats, triticale, lupins), and purchased compound feed. Primary data were collected from farm records and structured interviews (Table 1). The main differences between the two management systems included flock size, which decreased from 2,200 ewes and 150 rams under CG to 2,000 ewes and 120 rams under RG; the external feed share of the total diet, which declined from 37% to 25% for ewes and from 68% to 53% for lambs; the fertilizer application rate, which decreased from 120 to 38.5 kg NPK ha^-1 yr^-1; and diesel consumption, which declined from 31,900 to 11,700 L yr^-1, reflecting the shift from tillage-based feed production to integrated pasture management.

Each year, ewes produced an average of 1.34 lambs per ewe. A proportion of the female lambs born each year were retained internally as flock replacements to compensate for annual culling (approximately 21% of the ewe flock under CG and 19% under RG) and adult ewe mortality (1.5% yr^-1). Similarly, a small number of male lambs were retained annually as ram replacements. Lambs retained for internal replacement are not included in the “lambs sold” figures reported in Table 1; they enter the adult flock inventory and are recorded in the ewe or ram category when subsequently sold. All remaining lambs were sold for slaughter or as breeding stock, as shown in Table 1. The total live weight (LW) sold (77,250 kg yr^-1 pre-transition; 63,805 kg yr^-1 post-transition) reflects all animal sales across all categories and is the basis for the functional unit calculation. Animal performance parameters including individual body weight, mortality rate, and wool yield per head were consistent across both periods.

Secondary emission factors were sourced from reports published by the IPCC and FAO, with details provided in Supplementary Table S1 (IPCC, 2006; 2019; FAO, 2016; FAO, 2018). The net energy requirements calculated for rams, ewes, and lambs using the IPCC Tier 2 equations are reported in Supplementary Table S2. Portuguese datasets were used where available; otherwise, proxies from areas with similar climate, Spain, Italy, or Greece were applied based on crop type and agro-ecological comparability. Climate impacts were expressed as 100-year global warming potentials using IPCC Sixth Assessment Report characterization factors: CH₄ = 27.9 kg CO₂-eq kg^-1 and N₂O = 273 kg CO₂-eq kg^-1 (Forster et al., 2021).

2.3 Functional unit and allocation

Two functional units were used for our analysis: the farm, where climate impact was reported at farm-level, i.e., total GHG emissions associated with overall farm management (t CO₂-eq yr^-1); marketable products where climate impact was reported as the CF per kg of LW (kg CO₂-eq kg^-1 LW), consistent with ISO 14067 (ISO, 2018) and FAO LEAP guidelines for small-ruminant supply chains (FAO, 2016).

For GHG allocation, a consistent product mix was used for both periods (59% lamb LW, 39% ewe LW, 2% ram LW); wool production was estimated as 8,555 kg yr^-1 pre-transition; 7,717 kg yr^-1 post-transition. Results per kg hot carcass weight (HCW) are also provided in Supplementary Material (Supplementary Table S3) to facilitate comparison with literature data, as the latter is most commonly used for food analyses.

Environmental burdens were allocated between meat and wool using protein mass allocation (PMA) as the primary method, with protein contents set at 0.18 kg protein per kg LW and 0.65 kg protein per kg raw wool (Wiedemann et al., 2015). This allocation approach reflects the biological basis of feed conversion, remains stable under market price fluctuations, and aligns with FAO LEAP recommendations for small-ruminant LCA (FAO, 2016). The PMA factor for meat (AF_meat) was defined as the ratio of meat protein output to total protein output,where LW and W are the annual LW and raw wool sold and p_LW and p_W their respective protein contents (0.18 and 0.65 kg protein kg^-1). This yielded AF_meat = 0.714 pre-transition and 0.696 post-transition. The same factor was applied to both gross emissions and the soil C sequestration credit. For the post-transition period, the whole-farm sink of −1,812,980 kg CO₂-eq yr^-1, therefore, corresponded to a credit of 1,261,834 kg CO₂-eq yr^-1 allocated to meat (full workings in Supplementary Table S4). Economic allocation (EA) was applied as a sensitivity analysis; given the low market value of wool, EA attributed more than 98% of emissions to meat (Supplementary Table S4). An exploratory mixed variant (EA for emissions, PMA for accrual credits) is reported in Supplementary Table S5.

2.4 Enteric fermentation, manure and nutrient flows

Enteric CH₄ and manure-related CH₄ and N₂O were estimated using IPCC Tier 2 methodology (IPCC, 2006; IPCC, 2019). Forage digestibility was set at 60% for the post-transition period and 47% for the pre-transition period, reflecting improvement in pasture quality following the transition and consistent with IPCC and FAO LEAP ranges for sheep diets on Mediterranean grasslands (FAO, 2016; FAO, 2018). The pre-transition value of 47% reflects the low end of the range reported for degraded dryland pastures, consistent with field measurements of declining forage quality under CG pressure documented at the study site. Concentrate digestibility was 72% in both periods. Methane conversion factors (Y_m) followed IPCC defaults for sheep: 6.5% for adults and 4.5% for lambs. Full Tier 2 inputs by animal class and period are provided in Supplementary Tables S1, S2.

Grazed forage intake was determined by energy balance to meet net requirements for maintenance, activity, growth, lactation, pregnancy, and wool production (Supplementary Table S2), applying a 5% feed-waste factor (FAO, 2016). Nitrogen excretion was derived from intake and retention using standard values for LW and wool (FAO, 2016). Manure management followed actual farm practice: pasture deposition during grazing, short periods in unpaved yards, and solid storage. To avoid double-counting, pasture-deposited manure was attributed to agricultural soils and only yard and stored fractions counted under manure management emissions (IPCC, 2006; 2019). N₂O emissions from synthetic fertilizer application were estimated using IPCC default factors, with differences between periods directly reflecting the change in application rate shown in Table 1.

2.5 Soil carbon accrual estimation

Soil C accrual was quantified for the post-transition period 2018–2022, whereas C accrual pre-2018 was set as zero, as indicated by the low soil quality, including low C content (<2%), which suggested C loss rather than accrual before 2018. The losses of C in these areas were also confirmed by erosion-modelling calculations conducted for this area (Busico et al., 2023). Two complementary approaches were used to estimate the annual SOC accrual rate, and their consistency was assessed to support the reliability of the estimate used in the CF calculations. In the first approach (Approach 1), SOC stocks were measured at 0–30 cm depth across two representative sampling areas of approximately 4 ha each within the farm pasture in both 2018 and 2022. The annual accrual rate was calculated by dividing the total stock change by 4 years, assuming linear accrual over the monitoring period (Grilli et al., 2021).

In the second approach (Approach 2), SOC stocks measured in 2022 on the study farm were compared with stocks from a neighboring farm under conventional CG (Ishaq et al., 2025a), sampled in the same year using the same protocol. The difference in SOC stocks between the two farms was assumed to be the result of land management change in RG, with farm comparability in terms of soil type, parent material, and climate verified through prior site characterizations.

As the two approaches produced broadly similar estimates, we used data from Approach 1 to calculate the CO₂-eq removal associated with soil C accrual, as it draws directly on paired measurements from the same farm across the monitoring period and does not depend on assumptions about the comparability of an external reference site. Approach 2 is reported as a corroborating estimate.

C stocks were estimated using values of bulk density and total organic carbon (TOC) concentration measured in the top 30 cm of soil. Details on the sampled areas, design and number of replicates are provided in Grilli et al. (2021) and Ishaq et al. 2025a, b. In brief, bulk density was estimated using iron-steel cylinders that were inserted at 0–10, 10–20 and 20–30 cm. Soil samples were collected from the same profiles, air-dried, and passed through a 2-mm mesh sieve. Soil was then pulverized, and TOC content was measured using the Springer-Klee dichromate oxidation method, a modified ISO 14235:1998 method for the determination of organic carbon by sulfochromic oxidation, as outlined by Springer and Klee (1954).

3 Results

3.1 Whole-farm greenhouse gas emissions

Three emission categories were identified within the system boundary: farm operations, enteric fermentation, and manure management. Total farm emissions were 1,756.6 t CO₂-eq yr^-1 in the pre-transition period and 1,165.1 t CO₂-eq yr^-1 in the post-transition period, a reduction of 34% (Figure 2). Farm operations contributed 541.2 t CO₂-eq yr^-1 (31% of total) pre-transition and 221.1 t CO₂-eq yr^-1 (19%) post-transition, a reduction of 59%, driven primarily by lower purchased feed inputs, reduced diesel consumption, and less synthetic fertilizer use. Enteric fermentation was the dominant source in both periods, accounting for 1,083.9 t CO₂-eq yr^-1 (62%) pre-transition and 843 t CO₂-eq yr^-1 (72%) post-transition, a reduction of 22%. Manure management contributed 131.3 t CO₂-eq yr^-1 (7%) pre-transition and 101.2 t CO₂-eq yr^-1 (9%) post-transition, a reduction of 23%. Post-transition soil C accrual reached 494 t C yr^-1 (2.47 t C ha^-1 yr^-1 across 200 ha), equivalent to 1,813 t CO₂ yr^-1 removed, which exceeded total post-transition emissions of 1,165.1 t CO₂-eq yr^-1 by approximately 648 t CO₂-eq yr^-1, shifting the farm from a net source to a net sink (Figure 2).

FIGURE 2

3.2 Carbon footprint per unit of product

Moving from a farm-based analysis to a product-based analysis, the CF of sheep meat was 16.24 kg CO₂-eq kg^-1 LW in the pre-transition period and 12.71 kg CO₂-eq kg^-1 LW in the post-transition period, corresponding to a CF reduction of 22% (Table 2). Using the economic rather than mass allocation the CF values pre- and post-transition were 22.37 and 17.93 kg CO₂-eq kg^-1 LW. Results expressed per kg hot carcass weight are provided in Supplementary Table S3. For broader context comparison of the value of our results in terms of absolute sustainability, we compared the CF values of lamb meat produced in the studied farm from both periods to the distribution of 37 published CF estimates of lamb meat (Figure 3; Supplementary Table S7). The pre-transition CF value (16.24 kg CO₂-eq kg^-1 LW) falls at the 43rd percentile of the distribution of the 37 studies; the post-transition management further reduced the CF value (12.71 kg CO₂-eq kg^-1 LW, 32nd percentile of this distribution) increasing the difference with the average CF value reported in the literature for this product (18.3 kg CO₂-eq kg^-1 LW) (Figure 3).

FIGURE 3

3.3 Post-transition soil carbon accrual and net carbon balance of the farming system

Post-transition SOC accrual was estimated as 2.47 t C ha^-1 yr^-1 from paired farm measurements (Approach 1) and 2.92 t C ha^-1 yr^-1 using the comparison with the neighboring farm (Approach 2). The broad agreement between the two estimates supports the plausibility of the SOC accrual value, despite the different comparison frameworks. As anticipated, Approach 1 was used in all CF calculations. Scaling the observed C accrual rate over the 200 ha of the farm, this gives an annual C accrual of 494 t C yr^-1, equivalent to 1,813 t CO₂ yr^-1 removed. This exceeds the total post-transition farm emissions of 1,165.1 t CO₂-eq yr^-1, giving a net removal of approximately 648 t CO₂-eq yr^-1 (Figure 2). Allocating the overall C sink of −1,812,980 kg CO₂-eq yr^-1 at product-level using PMA (Section 2.3), the sequestration credit attributed to meat was 1,261,834 kg CO₂-eq yr^-1, against post-transition gross meat emissions of 810,886 kg CO₂-eq yr^-1, yielding a net removal of 450,948 kg CO₂-eq yr^-1. The resulting CF-OFFSET was −7.07 kg CO₂-eq kg^-1 LW under PMA and −9.97 kg CO₂-eq kg^-1 LW under EA (Table 2; detailed calculations in Supplementary Table S4). No temporal soil C measurements were available for the pre-transition period; CF-OFFSET therefore equals CF for that period. The sensitivity of these results to allocation method when sequestration credits are applied is explored in Supplementary Table S5.

4 Discussion

The management transition documented in this study shows that whole-farm changes in grazing strategy, input use, and land management can substantially alter both emission intensity and the C sink function of Mediterranean dryland sheep systems. Total farm GHG emissions fell by 34%, the gross CF of lamb decreased by 22% per kg LW, and measured soil C accrual was sufficient to shift the farm to a net C sink within 4 years. These outcomes mark the transition not merely as an emission-reduction measure but as an ecological intervention that restores degraded land and improves the system-level climate profile of livestock production (Ishaq et al., 2025b).

In degraded Mediterranean systems, pasture GHG emissions are associated with poor forage digestibility, high concentrate supplementation, and the loss of the C sink function that healthy grasslands can otherwise provide (Reyes-Palomo et al., 2024). Despite the recognized potential of RG, SOC dynamics are rarely incorporated into CF assessments of livestock systems and most LCA studies report gross emission intensities per unit of product without accounting for the C sink function of recovering pastures, which can substantially alter the net climate balance (Stanley et al., 2018; Forfora et al., 2024; Vagnoni et al., 2024).

We showed that the rationalization of the breeding system in the farm led to a 59% fall in farm operation emissions, reflecting decreased reliance on purchased concentrate feed, less diesel from reduced tillage, and lower fertilizer use. External input dependency is a primary determinant of farm-level CF in Mediterranean and European sheep systems (Pardo et al., 2023; Reyes-Palomo et al., 2024; Timpanaro and Foti, 2024), and by replacing purchased feed with on-farm forage production, the transition decoupled the farm from the upstream emission burden of industrial feed and fertilizer chains. This structural change is more durable than marginal improvements in feeding efficiency, which dominate mitigation strategies in intensive systems but leave the root cause of high operational emissions in degraded extensive systems untouched.

Although enteric fermentation remained the dominant source of GHG, as expected, its absolute reduction under RG (22%) is consistent with the smaller flock and the documented improvement in forage digestibility from 47% to 60% following the transition. Better pasture quality under rotational management therefore reduces emissions both directly, through lower methane conversion per unit of intake (Aguilera Sánchez and Molina Alcaide, 2021; Belanche et al., 2023), and indirectly, by reducing the need for high-emission concentrate supplementation.

The improved climate performance of the post-transition system was reflected in a decrease in the gross meat CF to 12.71 kg CO₂-eq kg^-1 LW, comparable to Mediterranean and semi-arid sheep systems CF values of 8.8–14.2 kg CO₂-eq kg^-1 LW for extensively managed sheep in southwestern Spain (Reyes-Palomo et al., 2024), and to CF values of 9.5–16.1 kg CO₂-eq kg^-1 LW found in semi-arid South Asian systems (Sarkar et al., 2024). The pre-transition value of 16.24 kg CO₂-eq kg^-1 LW sat at the higher end of these ranges, reflecting the higher dependency on external inputs of degraded pasture. Under EA, CF values increased by roughly one-third across both periods but produced a similar reduction of approximately 20%, confirming the observed change is robust to accounting assumptions (Wiedemann et al., 2015; Cottle and Cowie, 2016).

The soil C accrual rate of 2.47 t C ha^-1 yr^-1 is at the higher end of values reported for Mediterranean dryland grazing systems. Comparable studies report lower rates: approximately 0.9 t C ha^-1 yr^-1 in dehesa cattle systems in southwestern Spain (Reyes-Palomo et al., 2022), around 0.25 t C ha^-1 yr^-1 in Portuguese Montado agroforestry (Laporta et al., 2021), and 0.84 to 1.05 t C ha^-1 yr^-1 for improving degraded Portuguese grasslands nationally (Raposo et al., 2025). A study on Iberian dehesa woodlands confirmed that RG increases mineral-associated organic C and total SOC stocks through plant-microbial-nutrient pathways (Carrascosa et al., 2025). Outside the Mediterranean, Stanley et al. (2018) documented 3.6 t C ha^-1 yr^-1 over 4 years in adaptive multi-paddock beef systems on similarly degraded soils in the United States. The elevated rate here is consistent with the principle that SOC accumulation is fastest on severely depleted soils in early recovery (Conant et al., 2017; Bai and Cotrufo, 2022). The farm entered the transition with SOC below 2% and documented degradation across multiple soil quality indicators (Grilli et al., 2021; Ishaq et al., 2025b). Under CG, erosion rates were measured at 4.42 t ha^-1 yr^-1 at this site (Busico et al., 2023), a direct pathway of C loss that compounded SOC depletion. Reduced bare-soil exposure, increased vegetation cover, and lower surface runoff under the rotational scheme would have cut erosion-driven C loss alongside promoting in situ accumulation. Active pasture renovation, greater root biomass inputs, and reduced soil disturbance further supported C gains. SOC accumulation is inherently nonlinear, and rates decline as soil approach a new equilibrium; the 4-year window captures the early, fastest phase of this trajectory and should not be extrapolated as a long-term average. Consistency between the two estimation approaches, while acknowledging the limited sampling period and spatial coverage, supports the plausibility of the farm-scale estimate.

The post-transition soil C accrual of 494 t C yr^-1 (1,813 t CO₂ yr^-1) exceeded total farm emissions of 1,165.1 t CO₂-eq yr^-1, yielding a net farm C sink of 648 t CO₂-eq yr^-1 and negative CF-OFFSET values under both allocation methods. The CRCF establishes a framework for certifying precisely this type of outcome: documented, site-specific soil C gains alongside verifiable GHG reductions. The European Scientific Advisory Board on Climate Change (2026) recommended targeted transition support for high-emission, low-resilience farming systems and noted that Southern European regions face particularly high climate exposure, reinforcing the relevance of transitions of this kind in Mediterranean drylands. The monitoring approach used here, pairing temporal soil sampling with reference comparison, is broadly aligned with CRCF verification requirements, though full compliance would require longer time series, greater spatial coverage, and formal baseline accounting.

Several studies show that including SOC changes the interpretation of livestock system performance. In adaptive multi-paddock beef systems in the United States, Stanley et al. (2018) found the net CF shifted from 9.6 to −6.7 kg CO₂-eq kg^-1 HCW once SOC was credited. Batalla et al. (2015) found no net climate disadvantage in semi-extensive Spanish sheep farms after SOC inclusion. By contrast, Åby et al. (2024) reported marginal SOC contributions in Norwegian sheep systems under lower degradation pressure. The climate benefit of SOC is therefore most pronounced in systems recovering from degraded, low-carbon baselines, precisely the conditions characterizing Mediterranean dryland farms at desertification risk.

The transition also generated ecosystem service gains that reinforce the case beyond the C balance. Ishaq et al. (2025b) showed that the same management change produced quantifiable improvements across provisioning and regulating services under the Common International Classification of Ecosystem Services (CICES) framework, including water regulation, biodiversity, and fodder production, with SOC identified as the key mediating variable. Soil-level evidence from the same site documents gains in microbial biomass, aggregate stability, and water retention capacity (Ishaq et al., 2025a). These outcomes highlight a persistent gap in livestock LCA: the standard cradle-to-farm-gate boundary does not capture ecosystem service recovery, leading to a systematic undervaluation of well-managed extensive systems relative to more intensive ones (Horrillo et al., 2024; Vagnoni et al., 2024).

Despite the positive findings, we acknowledge some limitations of the present study. Because the intervention involved a whole-farm management transition, including simultaneous changes in grazing management, stocking density and input use, the before-and-after design evaluates the combined effect of the whole-farm transition and does not allow the effect of rotational grazing to be isolated from the other changes that occurred at the same time. The results should therefore be interpreted as the performance of the overall transition rather than of rotational grazing alone, and it is this combined transition that improved both farm and product sustainability. The soil C accrual rates reflect early-phase recovery from a severely degraded, low-carbon baseline and should be read as indicative of transition potential rather than long-term steady-state values; even at lower accrual rates, the rotational system still represents a meaningful improvement in the net GHG balance and in soil health and ecosystem services. Finally, there is currently no standardized method for crediting soil C within product-level LCA, which adds uncertainty to the CF-OFFSET values (Supplementary Tables S4 and S5).

5 Conclusion

This study provides field-based evidence that a whole-farm transition from continuous to rotational grazing can substantially improve the climate profile of a commercial sheep system in Mediterranean drylands, while simultaneously improving soil quality and ecosystem services. The findings carry four main implications. First, the emission intensity of extensive sheep-production systems is determined not only by the type of production system but also by dependence on external input and land condition. Second, farms transitioning from degraded continuous grazing can reduce gross GHG emissions and, when measured soil carbon accrual is credited, may shift their net carbon balance towards a sink. This finding challenges the assumption that extensive livestock systems are necessarily high emitting. However, the estimated net sink reflects early-phase soil carbon accrual measured over 4 years from a severely degraded baseline and depends on the assumptions used in soil carbon accounting. It should therefore be interpreted as evidence of transition potential rather than as a permanent steady-state condition. Third, the observed site-specific reductions in GHG emissions and increases in soil C accrual demonstrate the potential relevance of such management transitions for carbon farming strategies. Finally, the associated ecosystem-service benefits extend beyond those captured by CF accounting. Longer-term studies tracking SOC dynamics and ecosystem service provision beyond the initial transition phase are needed to characterize the long-term trajectory of farms undergoing similar management changes.

References

• 1

  ÅbyB. A.SamsonstuenS.BeaucheminK. A.AassL. (2024). Modelling of on-farm greenhouse gas emissions from dual-purpose meat and wool sheep production in different geographical regions of Norway. Livest. Sci.283, 105472. 10.1016/j.livsci.2024.105472

  • CrossRef
  • Google Scholar
  • View reference in article

• 2

  Aguilera SánchezJ. F.Molina AlcaideE. (2021). An estimation of enteric methane emissions from small ruminants in a mediterranean ecosystem. J. Agric. Environ. Sci.10 (2), 55–63. 10.15640/jaes.v10n2a6

  • CrossRef
  • Google Scholar
  • View reference in article

• 3

  AugustineD. J.DernerJ. D.Fernández-GiménezM. E.PorenskyL. M.WilmerH.BriskeD. D.et al (2020). Adaptive, multi-paddock rotational grazing management: a ranch-scale assessment of effects on vegetation and livestock performance in semiarid rangeland. Rangel. Ecol. Manag.73, 796–810. 10.1016/j.rama.2020.07.005

  • CrossRef
  • Google Scholar
  • View reference in article

• 4

  BaiY.CotrufoM. F. (2022). Grassland soil carbon sequestration: current understanding, challenges, and solutions. Science377, 603–608. 10.1126/science.abo2380

  • CrossRef
  • Google Scholar
  • View reference in article

• 5

  BatallaI.KnudsenM. T.MogensenL.del HierroÓ.PintoM.HermansenJ. E. (2015). Carbon footprint of milk from sheep farming systems in northern Spain including soil carbon sequestration in grasslands. J. Clean. Prod.104, 121–129. 10.1016/j.jclepro.2015.05.043

  • CrossRef
  • Google Scholar
  • View reference in article

• 6

  BelancheA.HristovA. N.van LingenH. J.DenmanS. E.KebreabE.SchwarmA.et al (2023). Prediction of enteric methane emissions by sheep using an intercontinental database. J. Clean. Prod.384, 135523. 10.1016/j.jclepro.2022.135523

  • CrossRef
  • Google Scholar
  • View reference in article

• 7

  BusicoG.GrilliE.CarvalhoS. C.MastrociccoM.CastaldiS. (2023). Assessing soil erosion susceptibility for past and future scenarios in semiarid mediterranean agroecosystems. Sustainability15 (17), 12992. 10.3390/su151712992

  • CrossRef
  • Google Scholar
  • View reference in article

• 8

  CapeloJ.Vila-ViçosaC. (2021). Vegetação do alentejo. A vegetação Port.1, 122–130.

  • Google Scholar
  • View reference in article

• 9

  CarrascosaA.MorenoG.CotrufoM. F.FradeC.RodrigoS.RoloV. (2025). “Improved management increases soil mineral-protected organic carbon storage via plant-microbial-nutrient mediation in semi-arid grasslands,”. EGUsphere. 10.5194/egusphere-2025-1711

  • CrossRef
  • Google Scholar
  • View reference in article

• 10

  CenteriC. (2022). Effects of grazing on water erosion, compaction and infiltration on grasslands. Hydrology9, 34. 10.3390/hydrology9020034

  • CrossRef
  • Google Scholar
  • View reference in article

• 11

  ConantR. T.CerriC. E.OsborneB. B.PaustianK. (2017). Grassland management impacts on soil carbon stocks: a new synthesis. Ecol. Appl.27 (2), 662–668. 10.1002/eap.1473

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 12

  CottleD. J.CowieA. L. (2016). Allocation of greenhouse gas production between wool and meat in the life cycle assessment of Australian sheep production. Int. J. Life Cycle Assess.21, 820–830. 10.1007/s11367-016-1054-4

  • CrossRef
  • Google Scholar
  • View reference in article

• 13

  de OtáloraX. D.EpeldeL.ArranzJ.GarbisuC.RuizR.MandalunizN. (2021). Regenerative rotational grazing management of dairy sheep increases springtime grass production and topsoil carbon storage. Ecol. Indic.125, 107484. 10.1016/j.ecolind.2021.107484

  • CrossRef
  • Google Scholar
  • View reference in article

• 14

  DelattreL.DeboliniM.PaoliJ. C.NapoleoneC.MouleryM.LeonelliL.et al (2020). Understanding the relationships between extensive livestock systems, land-cover changes, and CAP support in less-favored mediterranean areas. Land9, 518. 10.3390/land9120518

  • CrossRef
  • Google Scholar
  • View reference in article

• 15

  DondiniM.MartinM.De CamillisC.UwizeyeA.SoussanaJ.-F.RobinsonT.et al (2023). Global Assessment of Soil Carbon in Grasslands – from Current Stock Estimates to Sequestration Potential in FAO Anim. Prod. Health Pap (Rome: FAO), 187. 10.4060/cc3981en

  • CrossRef
  • Google Scholar
  • View reference in article

• 16

  European Parliament and Council (2024). Regulation (EU) 2024/3012 of the european parliament and of the council of 27 November 2024 establishing a union certification framework for permanent carbon removals, carbon farming and carbon storage in products. Off. J. Eur. Union L. Available online at: https://eur-lex.europa.eu/eli/reg/2024/3012/oj/eng (Accessed April 5, 2026).

  • Google Scholar
  • View reference in article

• 17

  European Scientific Advisory Board on Climate Change (2026). Climate Adaptation and Mitigation in the agri-food System. Luxembourg: Publications Office of the European Union. 10.2800/3706315

  • CrossRef
  • Google Scholar
  • View reference in article

• 18

  FAO (2016). Greenhouse Gas Emissions and Fossil Energy Use from Small Ruminant Supply Chains: Guidelines for Assessment. Rome: FAO. Available online at: https://www.fao.org/3/i6432e/i6432e.pdf (Accessed March 5, 2026).

  • Google Scholar
  • View reference in article

• 19

  FAO (2018). Nutrient Flows and Associated Environmental Impacts in Livestock Supply Chains: Guidelines for Assessment. Rome: FAO. Available online at: https://www.fao.org/3/i8798en/i8798en.pdf (Accessed March 5, 2026).

  • Google Scholar
  • View reference in article

• 20

  FAO (2025). “Greenhouse gas emissions from agrifood systems – global, regional and country trends, 2001–2023,” in FAOSTAT Anal. Brief. Ser.115. Rome. 10.4060/cd7300en

  • CrossRef
  • Google Scholar
  • View reference in article

• 21

  FerreiraC. S.Seifollahi-AghmiuniS.DestouniG.GhajarniaN.KalantariZ. (2022). Soil degradation in the european mediterranean region: processes, status and consequences. Sci. Tot. Env.805, 150106. 10.1016/j.scitotenv.2021.150106

  • CrossRef
  • Google Scholar
  • View reference in article

• 22

  ForforaN.AzuajeI.VivasK. A.VeraR. E.BritoA.VendittiR.et al (2024). Evaluating biomass sustainability: why below-ground carbon sequestration matters. J. Clean. Prod.439, 140677. 10.1016/j.jclepro.2024.140677

  • CrossRef
  • Google Scholar
  • View reference in article

• 23

  GasparP.EscribanoM.MesíasF. J.De LedesmaA. R.PulidoF. (2008). Sheep farms in the Spanish rangelands (Dehesas): typologies according to livestock management and economic indicators. Small Rumin. Res.74 (1–3), 52–63. 10.1016/j.smallrumres.2007.03.013

  • CrossRef
  • Google Scholar
  • View reference in article

• 24

  GrilliE.CarvalhoS. C.ChitiT.CoppolaE.D'AscoliR.La MantiaT.et al (2021). Critical range of soil organic carbon in southern Europe lands under desertification risk. J. Environ. Manage.287, 112285. 10.1016/j.jenvman.2021.112285

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 25

  HorrilloA.GasparP.Rodríguez-LedesmaA.EscribanoM. (2024). Assessment of greenhouse gas emissions and carbon sequestration in dairy goat farming systems in northern Extremadura, Spain. Animals14 (23), 3501. 10.3390/ani14233501

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 26

  IPCC (2006). Guidelines for National Greenhouse Gas Inventories: Agriculture, Forestry and Other Land Use. Geneva: IPCC. Available online at: https://www.ipcc-nggip.iges.or.jp/public/2006gl/ (Accessed February 7, 2026).

  • Google Scholar
  • View reference in article

• 27

  IPCC (2019). 2019 Refinement to the 2006 IPCC Guidelines for National Greenhouse Gas Inventories, 4. Geneva: AFOLU. IPCC.

  • Google Scholar
  • View reference in article

• 28

  ForsterP.StorelvmoT.ArmourK.CollinsW.DufresneJ.-L.FrameD.et al (2021). “Climate change 2021: the physical science basis,” in Contribution of Working Group I to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change. Editor Masson-DelmotteV. (Cambridge: Cambridge University Press). Available online at: https://www.ipcc.ch/report/ar6/wg1/chapter/chapter-7/#7.6 (Accessed February 8, 2026).

  • Google Scholar
  • View reference in article

• 29

  IshaqH. K.GrilliE.D'AscoliR.MastrociccoM.RutiglianoA. F.MarzaioliR.et al (2025a). Soil quality under rotational and conventional grazing in mediterranean areas at desertification risk. J. Environ. Manage.373, 123822. 10.1016/j.jenvman.2024.123822

  • CrossRef
  • Google Scholar
  • View reference in article

• 30

  IshaqH. K.GrilliE.BusicoG.MastrociccoM.RutiglianoA. F.BijlM.et al (2025b). Enhancing ecosystem services by introducing multifunctional land use management in a semiarid pastoral system in Portugal. Total Environ. Adv.16, 200138. 10.1016/j.teadva.2025.200138

  • CrossRef
  • Google Scholar
  • View reference in article

• 31

  ISO (2006a). ISO 14040: Environmental Management — Life Cycle Assessment — Principles and Framework. Geneva: ISO.

  • Google Scholar
  • View reference in article

• 32

  ISO (2006b). ISO 14044: Environmental Management — Life Cycle Assessment — Requirements and Guidelines. Geneva: ISO.

  • Google Scholar
  • View reference in article

• 33

  ISO (2018). ISO 14067: Greenhouse Gases — Carbon Footprint of Products — Requirements and Guidelines for Quantification. Geneva: ISO.

  • Google Scholar
  • View reference in article

• 34

  KairisO.KaravitisC.SalvatiL.KounalakiA.KosmasK. (2015). Exploring the impact of overgrazing on soil erosion and land degradation in a dry mediterranean agro-forest landscape (crete, Greece). Arid. Land Res. Manag.29, 360–374. 10.1080/15324982.2014.968691

  • CrossRef
  • Google Scholar
  • View reference in article

• 35

  KosmasC.DetsisV.KaramesoutiM.KounalakiK.VassiliouP.SalvatiL. (2015). Exploring the long-term impact of grazing management on land degradation in the socio-ecological system of the asteroussia Mountains, Greece. Land4, 541–559. 10.3390/land4030541

  • CrossRef
  • Google Scholar
  • View reference in article

• 36

  LaportaL.DomingosT.Marta-PedrosoC. (2021). It's a keeper: valuing the carbon storage service of agroforestry ecosystems in the context of CAP eco-schemes. Land Use Policy109, 105712. 10.1016/j.landusepol.2021.105712

  • CrossRef
  • Google Scholar
  • View reference in article

• 37

  Life desert-adapt (2026). Preparing Desertification Areas for Increased Climate Change: LIFE16 CCA/IT/000011. Available online at: https://www.desert-adapt.it/ (Accessed July 22, 2026).

  • Google Scholar
  • View reference in article

• 38

  OggioniS. D.Ochoa-HuesoR.PecoB. (2020). Livestock grazing abandonment reduces soil microbial activity and carbon storage in a mediterranean dehesa. Appl. Soil Ecol.153, 103588. 10.1016/j.apsoil.2020.103588

  • CrossRef
  • Google Scholar
  • View reference in article

• 39

  PardoG.CasasR.del PradoA.ManzanoP. (2023). Carbon footprint of transhumant sheep farms: accounting for natural baseline emissions in mediterranean systems. Int. J. Life Cycle Assess.29, 2184–2199. 10.1007/s11367-023-02135-3

  • CrossRef
  • Google Scholar
  • View reference in article

• 40

  RaposoM.CanaveiraP.DomingosT. (2025). Estimating soil carbon sequestration potential in Portuguese agricultural soils through land-management and land-use changes. Sustainability17 (3), 1223. 10.3390/su17031223

  • CrossRef
  • Google Scholar
  • View reference in article

• 41

  Reyes-PalomoC.AguileraE.LlorenteM.Díaz-GaonaC.MorenoG.Rodríguez-EstévezV. (2022). Carbon sequestration offsets a large share of GHG emissions in dehesa cattle production. J. Clean. Prod.358, 131918. 10.1016/j.jclepro.2022.131918

  • CrossRef
  • Google Scholar
  • View reference in article

• 42

  Reyes-PalomoC.Díaz-GaonaC.Sanz-FernándezS.Muñoz-CobosI.AguileraE.Rodríguez-EstévezV. (2024). Carbon footprint of an extensively raised, low-productivity sheep population. Agriculture14 (8), 1304. 10.3390/agriculture14081304

  • CrossRef
  • Google Scholar
  • View reference in article

• 43

  RoxoM. J.CerqueiraH. M.Calvo‐CasesA. (2023). The role of land use and land cover changes in triggering soil losses in the SE alentejo, Portugal. Land Degrad. Dev.34 (18), 5954–5968. 10.1002/ldr.4889

  • CrossRef
  • Google Scholar
  • View reference in article

• 44

  SabaterS.ElosegiA.FeioM. J.GómezR.GraçaM. A.MuñozI.et al (2022). “The iberian Rivers,” in Rivers of Europe. Editors TocknerK.UehlingerU.RobinsonC. T.2nd ed (Elsevier), 181–224. 10.1016/B978-0-08-102612-0.00004-3

  • CrossRef
  • Google Scholar
  • View reference in article

• 45

  SarkarS.LalB.GautamP.BhattR. S.SahooA. (2024). Comparing carbon footprints of sheep farming systems in semi-arid regions of India: a life cycle assessment study. PLoS One19 (1), e0292066. 10.1371/journal.pone.0292066

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 46

  SpringerU.KleeJ. (1954). Prüfung der Leistungsfähigkeit von einigen wichtigeren Verfahren zur Bestimmung des Kohlenstoffs mittels Chromschwefelsäure sowie Vorschlag einer neuen Schnellmethode. Z. Pflanzenernähr. Düng. Bodenk.64, 1–26. 10.1002/jpln.19540640102

  • CrossRef
  • Google Scholar
  • View reference in article

• 47

  StanleyP. L.RowntreeJ. E.BeedeD. K.DeLongeM. S.HammM. W. (2018). Impacts of soil carbon sequestration on life cycle greenhouse gas emissions in midwestern USA beef finishing systems. Agric. Syst.162, 249–258. 10.1016/j.agsy.2018.02.003

  • CrossRef
  • Google Scholar
  • View reference in article

• 48

  TeutscherováN.VázquezE.SoteloM.VillegasD.VelásquezN.BaqueroD.et al (2021). Intensive short-duration rotational grazing is associated with improved soil quality within one year after establishment in Colombia. Appl. Soil Ecol.159, 103835. 10.1016/j.apsoil.2020.103835

  • CrossRef
  • Google Scholar
  • View reference in article

• 49

  TimpanaroG.FotiV. T. (2024). The sustainability of small-scale sheep and goat farming in a semi-arid mediterranean environment. J. Sustain. Agric. Environ.3 (3), e12111. 10.1002/sae2.12111

  • CrossRef
  • Google Scholar
  • View reference in article

• 50

  VagnoniE.ArcaP.DecandiaM.MolleG.SerraG.SauP.et al (2024). Looking for the ecological transition of mediterranean small ruminant sector: characterization and main drivers of environmental performance of the Sardinian sheep farming systems. Clean. Environ. Syst.14, 100214. 10.1016/j.cesys.2024.100214

  • CrossRef
  • Google Scholar
  • View reference in article

• 51

  WiedemannS. G.LedgardS. F.HenryB. K.YanM. J.MaoN.RussellS. J. (2015). Application of life cycle assessment to sheep production systems: investigating co-production of wool and meat using case studies from major global producers. Int. J. Life Cycle Assess.20, 463–476. 10.1007/s11367-015-0849-z

  • CrossRef
  • Google Scholar
  • View reference in article

• 52

  ZhangR.TianD.ChenH. Y.SeabloomE. W.HanG.WangS.et al (2022). Biodiversity alleviates the decrease of grassland multi-functionality under grazing disturbance: a global meta-analysis. Glob. Ecol. Biogeogr.31, 155–167. 10.1111/geb.13408

  • CrossRef
  • Google Scholar
  • View reference in article