Environmental Performance of a Mixed Crop–Dairy Cattle Farm in Alexandria (Romania)

Abstract

Agricultural specialization has increased considerably in Europe over the last decades, leading to the separation of crop and livestock production at both farm and regional levels. Such a transformation is often associated with higher environmental burdens due to excessive reliance on exogenous inputs and manure management issues. Reconnecting crop and livestock production via mixed farming systems (MFSs) could improve circularity and resilience, leading to reduced environmental impacts. The objective of this study was to evaluate the life cycle environmental performance of a commercial mixed crop–dairy cattle farm in Romania and to compare it against the corresponding specialized systems. The evaluation covered both dairy cattle production (milk and meat) and cash crops. Overall, the results show that the coupled system improves environmental performance by reducing the over-reliance on high-impact inputs like synthetic fertilizers and exogenous feed. The carbon footprint for the milk production of the studied system (1.17 kg CO₂ eq.) per kg of fat- and protein-corrected milk (FPCM) was 10% lower than the mean value of common intensive milk production systems. The eutrophication impacts (2.52 × 10⁻⁴ kg P eq and 2.67 × 10⁻⁴ kg N eq./kg of FPCM) presented values of one order of magnitude less than their specialized counterparts. However, the impacts of the studied MFS, albeit lower than those for comparable specialized systems, still remain relatively high. In particular, methane emissions from enteric fermentation (0.54 kg CO₂ eq./kg FPCM) were a major contributor to the carbon footprint. This highlighted the need to address the elevated emissions from enteric fermentation with better feed management, as well as improving and reinforcing the system’s self-sufficiency.

1. Introduction

The environmental impacts of agriculture, which include pollution, overexploitation of natural resources, biodiversity decline, and greenhouse gas (GHG) emissions, have been aggravated by the specialization and intensification of agricultural systems [1,2]. In Europe, nearly 60% of farms were specialized in crops and 22% in livestock in 2020 [3]. The lack of interaction between livestock and crop production implies heavy reliance on external inputs and excess manure in intensive livestock farms, resulting in declining soil organic matter and fertility, as well as significant water pollution [3]. These serious issues have led to growing interest among policy makers and researchers in mixed farming systems (MFSs) as a possible alternative to limit the negative impacts of specialization. MFSs are characterized by the coexistence of cash and feed crops and livestock activities on the same farm, resulting in different subsystems with shared inputs and benefits [4].

The reported positive effects of reconnecting livestock and crop production include better manure recycling, sustainable intensification, reducing feed imports and improving resilience and biodiversity [5,6,7]. However, such benefits have mainly been evidenced at the theoretical level and further information and knowledge about the environmental impacts are needed at the practical level [8]. There is a wide diversity of MFSs based on management practices, integration levels, and regional models [9]. These multifunctional systems have multiple production outputs, in addition to different synergies and interactions between farm components and, therefore, the assessment of the environmental impacts is challenging [10].

Life cycle assessment (LCA) is a standardized methodology [11,12] which can address the complexity and diversity of agricultural systems, identify critical environmental issues, and select the best alternatives. LCA is generally recommended and applied to evaluate the environmental performance of agricultural systems [13,14] and has been extensively used to assess crops, livestock systems, and organic farming [15,16]. However, the number of LCAs focused on MFSs is very limited compared to other agricultural systems [17]. Moreover, most such studies used data from statistical databases [4,18,19] or experimental sites [20,21].

Overall, these previous LCA studies have reported that MFSs outperform specialized systems in terms of environmental friendliness. Marton et al. [4] revealed the potential of MFSs to minimize environmental impacts compared to specialized dairy farming in Switzerland, by using system expansion to account for the advantages and downsides of manure application out of the bounds of the dairy system. Nonetheless, Tendall and Gaillard [18] showed that the MFS footprints in Switzerland are also likely to increment in the future climate and that agricultural accommodation is, therefore, needed. They emphasized that more integrative and regionally differentiated policies will be required, as each region is expected to react differently. Parajuli et al. [19] demonstrated that the integration of MFSs with a green biorefinery and a biogas conversion facility increased circularity and resulted in better environmental impacts compared to a conventional MFS in Denmark. Paramesh et al. [20] concluded that the expansion of MFSs with fishery and poultry components and the diversification of crops is crucial to compensate for the ecological imbalances resulting from continuous rice cultivation in India. In Vogel et al. [21], two experimental crop–livestock systems with more rotations (including beef cattle with various combinations of rice and soybean crops) achieved a better environmental performance compared to the traditional crop–livestock system in Brazil.

In contrast, Veysset et al. [22] analyzed real case farms in France and reported that specialized systems had more efficient use of inputs and, overall, performed better than MFSs, positioning organic farming as a prototype MFS meeting the core principles of agroecology. However, those authors only focused on a single MFS product (beef production) and did not consider the comparability limitations between farms when assessing impacts, such as animal and crop conditions (confinement, rainfed, etc.). More recently, studies in China by Ma et al. [7] at a regional level and by Xu et al. [23] at a farm level used data from farm surveys for the two mixed farming outputs (livestock and crop products). These authors reported a decrease in GHG emissions by increasing cooperation and integration between crops and livestock farms, once again highlighting the potential benefits of reconnecting these systems.

The main objectives of this study are to determine the life cycle environmental impacts of a commercial mixed crop–dairy cattle farm in Romania, to identify the critical aspects and to evaluate the potential footprint improvement of the coupled system compared to specialized systems. Furthermore, the work aims to contribute to the understanding of how reconnecting crop and livestock production and hence boosting circularity, in terms of feed self-sufficiency and manure recycling, within an intensive livestock production system could lead to a better environmental performance. This is particularly relevant in Romania, since it is the European country which experienced the most significant reduction in the number of mixed farms between 2005 and 2020 [3]. The drastic decline of such systems makes the collection of data from real cases more complex, especially for data-intensive methodologies such as LCA. As far as the authors know, this is the first LCA of a closely monitored mixed farming system in Romania, which includes the evaluation of eight feed crops, three cash crops, and two livestock products (milk and meat). This work provides valuable information for policy making, aiming to ameliorate the sustainability of dairy farming by coupling crop and livestock production at the farm level.

2. Materials and Methods

2.1. Case Study

A mixed intensive crop–dairy cattle farm located in Moșteni village, in the north-eastern part of the Teleorman county in the Alexandria region (Romania, Figure A1), was monitored for the three-year period 2021–2023, in order to collect the input records for the LCA study. The region has a humid, subtropical climate characterized by warm and humid summers and winters with moderate temperatures. The farm covers a total area of 675 ha, which is used for growing rainfed crops (660 ha) and to keep animals confined (i.e., they stay in the dairy farm or do not graze at any time of the year). The main characteristics of the studied farm, which is categorized as fully coupled crop–livestock production [23], are summarized in

Table 1. Main features of the mixed crop–dairy cattle farm studied.

Parameter	Unit	Quantity
Farm area	ha	675
Animals
Stable area	m2	2240
Milking cows	heads	150
Heifers and primiparous	heads	40
Young stock	heads	10
Duration of lactation	days	305
Interval between births	months	14
Average weaning age	months	3
First birth age	months	29–30
Number of lactations	lactations	6–7
Dismissal age	years	9–10
Crops	ha	660
Corn silage	ha	30
Wheat	ha	234.2
Barley	ha	47.3
Corn kernels	ha	60.7
Oat	ha	1.7
Rapeseed	ha	130.4
Sunflower	ha	114.5
Alfalfa	ha	41.2

. The farm is representative of mixed farms in Romania, having an average size, both in terms of the number of animals and cultivated area [3]. This farm was chosen since it had the specific characteristics required to comply with the MFS definition [4,24,25], which include circularity features like high feed self-sufficiency (85%) and return of manure to the cropland (100%), as well as a proportion of cash crops (crops sold).

Cattle production includes milk and live weight for meat production. Crops (corn, wheat, barley, oat, rapeseed, sunflower, and alfalfa) are grown annually to provide feed for the animals and also cash crops. The latter corresponds to 69% of the wheat production, 49% of the corn kernels, and 55% of the barley. The harvested fodder is stored at the farm in a warehouse with a concrete structure and floor and natural ventilation. The straw produced in the cultivated area is used as bedding material for animal housing. It should be noted that local crops provide about 85% of the animal feed, so only 15% is imported.

All the manure from the livestock activity is applied as fertilizer for the crops. Manure is separated into liquid and solid forms prior to application to the soil. The solid part is stored in a manure platform with a capacity of 1200 t, and the liquid part in a 4600 m³ lagoon lined with a 2 mm thick impermeable membrane. Manure from livestock production is transported and spread on the field crops every year, with a storage period ranging from 180 to 365 days. The manure produced provides the following proportions of the total fertilizer nutrients applied to field crops: 15% of nitrogen (N), 20% of phosphorus (P₂O₅), and 50% of potassium (K₂O).

2.2. Life Cycle Assessment

The LCA was performed following the ISO standards for LCA, according to the guidelines and requirements [11,12]. An attributional approach was selected to determine the environmental performance of the MFS, following the four steps of the standardized LCA framework, (i) goal and scope definition; (ii) life cycle inventory (LCI); (iii) life cycle impact assessment (LCIA); and (iv) interpretation of results.

2.2.1. Goal and Scope Definition

The goal of the study was to assess the environmental footprint of an MFS in the region of Alexandria (Romania) and to compare the coupled system footprint with the corresponding specialized systems. The scope was the ‘cradle-to-gate’ production of milk, meat, and cash crops. The environmental performance of these products sold by the farm was compared with their counterparts in previous LCA studies of specialized systems (Section 4). Figure 1 shows the flow diagram of the processes incorporated in the LCA and the system boundaries.

The functional unit (FU) for milk production was 1 kg of fat- and protein-corrected milk (FPCM). FPCM is calculated in accordance with the International Dairy Federation’s guide for the LCA of dairy products [26]. Economic allocation has been used to allocate impacts between milk and its co-product, liveweight, for meat production. The economic allocation, based on generated incomes, was previously evaluated as being adequate for bovine milk production [27,28]. For crops, the FUs selected were 1 t of mass produced and 1 ha of cultivated land.

The different stages (or processes) and relative flows were grouped into the following four subsystems: crop management, animal housing, feed management, and manure management. The details of the stages and flows of each subsystem are provided below:

Crop management

-

Seeds: this involves the production of seeds for sowing.

-

Synthetic fertilizers: this includes the production, transport, and packaging of synthetic fertilizers and the emissions of the field application of the nutrients (N, P₂O₅, and K₂O). The following emissions are accounted for: emissions to air of ammonia (NH₃), nitrogen oxides (NO_x), and indirect and direct nitrous oxides (N₂O), as well as emissions to water of nitrates (NO₃⁻) and phosphates (PO₄³⁻).

-

Manure application: this stage concerns the pollution caused by the transport and the field emissions of the application of the manure. The above-mentioned emissions (NH₃, NO_x, N₂O, NO₃⁻, and PO₄³⁻) were also determined for this stage.

-

Pesticides: this accounts for the manufacturing, transport, packaging, and application of fungicides, herbicides, and insecticides to the cropping area.

-

Machinery: this considers the fabrication, conservation, and use of machinery for sowing seeds, spreading fertilizers (synthetic and manure), soil management, and harvesting. Fertilizers and pesticides were applied to the field using bottom-spreading machinery attached to a four-wheel tractor. Fuel consumption and associated atmospheric emissions were also considered.

Animal housing

-

This includes the production and use of inputs needed to maintain the confined animals in the barn. These include water, electricity, and the management and transport of bedding materials (straw), as well as fuel consumption and the resulting emissions.

-

Methane (CH₄) emissions from enteric fermentation were included in this stage.

Feed management

-

This considers the production and transport of exogenous feed (soybean meal and mineral salt) and feed (produced on the farm) transport.

-

The fuel consumption and emissions associated with this stage were accounted for.

Manure management

-

This covers the production and consumption of electricity and fuel for pumping and processing the manure produced.

-

Manure emissions of CH₄ and N₂O (direct and indirect) and the relative atmospheric emissions from diesel fuel were calculated.

2.2.2. Life Cycle Inventory

The inputs and outputs included in the life cycle stages of the MFS’s subsystems (crop management, animal housing, feed management, and manure management) are presented in

Table 2. Main inputs and outputs for the stages of each subsystem included in the life cycle of the mixed crop–dairy cattle farm.

Subsystem
Inputs/Outputs	Unit	Corn Silage	Wheat	Barley	Corn Kernels	Oat	Rapeseed	Sunflower	Alfalfa
Crop management
Seeds	kg ha−1	20	250	180	22	125	2.20	4	25
Synthetic fertilizers
N	kg ha−1	112.60	129.33	128.16	112	125	93.80	95	110
P2O5	kg ha−1	3	13.70	10.80	12.50	11	25	17.50	1.50
K2O	kg ha−1	5.50	16.40	22.30	16.50	31.20	37	50	6.50
Manure application
N	kg ha−1	16.40	28	24.60	41.80	28	8.20	8.20	0
P2O5	kg ha−1	4	2.33	6	5.69	2.33	2	2	0
K2O	kg ha−1	10	47.90	15	56.30	47.90	5	5	0
Emissions to air
Ammonia (NH3)	kg ha−1	8.96	12.09	11.36	13.93	11.87	6.39	6.45	5.59
Nitrogen oxides (NOx)	kg ha−1	5.16	6.29	6.11	6.15	6.12	4.08	4.13	4.4
Nitrous oxides (N2O)	kg ha−1	1.14	1.39	1.35	1.38	1.36	0.89	0.90	0.95
Emissions to water
Nitrate (NO3−)	kg ha−1	8.35	9.95	9.74	9.30	9.65	6.75	6.83	7.54
Phosphate (PO43−)	kg ha−1	0.35	0.80	0.84	0.91	0.67	1.35	0.98	0.08
Pesticides	kg ha−1	2.21	2.21	0.90	2.21	2.21	0.90	3.25	0.70
Machinery
Time	h ha−1	5	5.17	6	7	5.15	4.67	4.87	3.90
Diesel	kg ha−1	33.50	34.39	39.94	46.59	34.28	31.06	32.39	26
Yield
Crop production	t ha−1	49.75	6	6	10	3.68	3.25	2.73	46
Straw production	t ha−1	0	4.82	4.54	13	2.93	10.37	10.42	0
Subsystem
Inputs/Outputs	Unit	Quantity
Animal housing
Electricity	kWh year−1	91,800
Water	m3 year−1	4461
Wastewater	m3 year−1	4583
Bedding material	t year−1	365.04
Diesel	kg year−1	200
Enteric fermentation
CH4 emissions	t year−1	21.48
Feed management
Soybean meal	t year−1	124.10
Salt	t year−1	1.60
Diesel	t year−1	5.87
Manure management
Milking cows’ manure	t year−1	2463.75
Heifers’ manure	t year−1	657
Young stock manure	t year−1	24.30
Electricity	kWh year−1	10,200
Diesel	kg year−1	485
CH4 emissions	kg year−1	1031.30
N2O emissions	kg year−1	381.89
Production	Unit	Quantity
Milk	kg FPCM year−1	1,376,337.54
Meat for live weight	kglw year−1	500

. Most of the foreground (primary) data were collected directly from the farmer and some was collected from the literature. The data for the background activities were extracted from ecoinvent v.3.9.1 [29].

Table A1. Data sources for the stages of each subsystem included in the mixed crop–dairy cattle farm.

Subsystem/Stage	Inputs/Outputs/Parameters	Data Sources
Productivity and farm description		Farmer
Crop management
Seeds	Doses	Farmer
	Production and transport	[29]
Synthetic fertilizers	Manufacturing and transport	[29]
	Doses, type, and application	Farmer
	Field emissions	[31,32,33,34,35]
Manure application	Doses, type, and application	Farmer
	Field emissions	[31,32,34,35]
Machinery	Equipment	Farmer
	Machinery use	Farmer, [30]
	Diesel emissions	[29]
Plant Protection Products	Manufacturing and transport	[29]
	Doses, type, and application	Farmer
Animal housing
Water and energy (electricity, diesel)	Amount	Farmer
	Manufacturing and emissions	[29]
Bedding material	Amount of straw	Farmer
	Amount of energy, transport, and emissions	Farmer, [29]
Enteric fermentation	Feed ration and characteristics	Farmer, [36]
	Gross energy intake (Equation)	[36,37]
	Methane conversion factor (Equation)	[58]
	CH4 emissions	[31]
Feed management	Amount	Farmer
	Exogenous feed production	[29]
	Amount of energy, transport, and emissions	Farmer, [29]
Manure management	Amount	Farmer
	Feed ration and characteristics, gross energy intake and annual N excretion rates	Farmer, [31,36,37]
	Amount of energy, transport, and emissions	Farmer, [29,31]

(Appendix A) shows the data sources used to develop the LCI. A detailed description of the foreground and background data is provided below:

• Foreground data:

-

Farm characteristics, components, resources, and field practices, which included herd and crop structure, management practices, production, machinery, water, energy consumed, chemical inputs, and feed were obtained directly from farm records during periodic visits in 2021–2023. Data on machinery and equipment (time, type, diesel, etc.) were supplemented with technical recommendations provided by Audsley et al. [30] to make the calculations on machinery use. The diesel emissions were derived from ecoinvent.

-

Emissions from synthetic fertilizers and manure application were calculated from (i) IPCC guidelines [31] for the air emissions of N₂O and NH₃ from manure and the indirect emissions of N₂O from synthetic fertilizers; (ii) EMEP/EEA guidelines [32] for the air emissions of NO_x from manure, as well as NH₃ and NO_x from synthetic fertilizers; (iii) data from Menegat et al. [33] (a regional estimate per country) for the direct emissions to the air of N₂O from synthetic fertilizers; and (iv) methodologies proposed by Wang et al. [34] and Nemecek and Kagi. [35] for NO₃⁻ and PO₄³⁻ leaching to groundwater for manure and synthetic fertilizers.

-

The IPCC guidelines [31] were followed to calculate CH₄ from enteric fermentation and manure application (Tier2 methodology), as well as to calculate N₂O emissions from the latter subsystem. Information on the feed ration for each animal sub-category is shown in

Table A2. Feed ration in the mixed crop–dairy cattle farm.

Category *	Dairy Cattle (kg/Head/Day)	Primiparous and Heifers (kg/Head/Day)	Youth (3–9 Months) (kg/Head/Day)
Barley straw	1	1.5	1
Alfalfa hay	3	3.5	2
Corn silage	22	15	7
Corn kernels	2	1.5	0.5
Barley kernels	1.5	1.5	0.3
Rapeseed meal	1.5	1	0.3
Wheat bran	2	0	_
Soybean meal	2	1	_

* For dairy cattle and primiparous and heifers, 5 kg/head/day of green mass are added in summer and 5 kg/head/day of beer wort are added in winter.

(Appendix A). The main characteristics of the feed ration were retrieved from previous work by the authors [36]. The methodology used to estimate the energy intake was also retrieved from the cited work, which was based on the equation of Stoica et al. [37]. The sources used to calculate the main parameters for estimating the emissions are listed in Table A1 (Appendix A). The average annual population for each subcategory was calculated considering the average herd composition for 2021–2023, as follows: milking cows (130), heifers and primiparous cows (40), and young cattle (30).

• Background data:

Data for the following stages and the associated emissions were extracted from the ecoinvent v.3.9.1 database [29]:

-

Production of seeds for sowing.

-

Manufacturing processes and transport of agricultural machinery.

-

Manufacturing, packaging, and transport of synthetic fertilizers and pesticides.

-

Water and energy production (electricity and fuel).

-

Exogenous feed production (soybean meal and salt).

2.2.3. Life Cycle Impact Assessment

For the impact assessment, midpoint impact categories are widely used and recommended in LCAs applied to agricultural production [18,23]. Since the analysis addressed both the crop and the livestock systems, we focused on impact categories relevant to both systems and directly correlated this with the estimated emissions. In a broader context, other indicators should be considered but, here, a selection of impact categories relevant to the case study was made based on multifunctionality, in order to simplify the output for decision support and problem solving [18,38]. Accordingly, the following impact categories were selected: carbon footprint (CF, kg CO₂ eq), freshwater eutrophication (FE, kg P eq), terrestrial acidification (TA, kg SO₂ eq), marine eutrophication (ME, kg N eq), and freshwater ecotoxicity (FWE, kg 1,4-DCB). The ReciPe 2016 midpoint (H) method [39] was selected for impact calculations, which were performed with the software openLCA v.1.11.0.

3. Results

The first results presented in this section are those for the impacts of the whole mixed farming system and then for the subsystems of the farm (crop management, animal housing, feed management, and manure management).

3.1. Impacts of the Mixed Farming System

The MFS impacts (livestock products and crops sold) with and without allocation between livestock products (milk and meat) are shown in

Table 3. Environmental impacts of the mixed farming system.

(a) Impacts of Livestock Products (per kg FPCM FU)
Impact Categories	Unit	No Allocation (All Impacts for Milk)	Economic Allocation
		Milk	Milk	Meat
CF	kg CO2 eq	1.19	1.17	1.19 × 10−2
FE	kg P eq	2.52 × 10−4	2.49 × 10−4	2.52 × 10−6
ME	kg N eq	2.67 × 10−4	2.65 × 10−4	2.67 × 10−6
TA	kg SO2 eq	4.09 × 10−3	4.05 × 10−3	4.09 × 10−5
FWE	kg 1,4-DCB	1.12 × 10−2	1.11 × 10−2	1.12 × 10−4
(b) Impacts of Crop Products Sold (per t FU)
Impact Categories	Unit	Wheat	Barley	Corn Kernels
CF	kg CO2 eq	3.71 × 102	3.53 × 102	1.93 × 102
FE	kg P eq	2.11 × 10−1	1.95 × 10−1	1.19 × 10−1
ME	kg N eq	2.01 × 10−1	1.94 × 10−1	7.96 × 10−2
TA	kg SO2 eq	5.30	4.90 × 10−3	3.34
FWE	kg 1,4-DCB	7.68	7.18	3.85

. Relevant differences were not observed between the impact values of milk without allocation (100% to milk) and after economic allocation (Table 3a), since meat production in the MFS studied is considered insignificant (Table 2). The percentual contribution of the subsystems to the impacts of livestock production is presented in Figure 2. This shows that the CF impact category was mainly influenced by animal housing (45.6%), due to high methane emissions from enteric fermentation. However, the crop management subsystem generated the highest burdens in the FE, TA, and FWE impact categories (up to 78% in FE and 87% in TA). These were mostly attributed to the production and application of synthetic fertilizers (Figure 3), as highlighted in Section 3.2.1. The only category dominated (56%) by feed management was ME, mostly caused by the exogenous soybean meal but, once again, it was closely followed by the crop management subsystem (43%), due to the input of synthetic fertilizers. The high burdens caused by the industrial production and field application of synthetic fertilizers are often identified as one of the primary pollutants in agricultural activities, particularly nitrogen fertilizers [40,41,42]. Overall, the external nutrients imported to the MFS had the most notable negative environmental impact.

The remaining impacts of the mixed farm system were generated by the crops sold (wheat, corn kernels, and barley), shown in Table 3b. Wheat cultivation had the highest CF per mass produced and the topmost impacts in the other studied categories, followed by barley and corn kernels (Table 3b), as indicated in Section 3.1 (Figure A2).

3.2. Contribution Analysis

3.2.1. Crop Management

The average contribution of the stages of the cultivated crops to the selected impact categories in the studied MFS are shown in Figure 3. N₂O and CO₂ emissions to the air caused by the fabrication and use of synthetic fertilizers were the main pollution source in the CF impact category, with an average contribution of 84%, followed by machinery, albeit with a much lower contribution (8%). FE was mostly caused by phosphorus emissions to water from synthetic fertilizers and manure application, emitting an average of 78% and 14% of total phosphorus (kg P eq), respectively. Leaching of NO₃⁻ to groundwater was the main pollution source in the ME category, stemming from synthetic fertilizers and seed production, with average shares of 71% and 23%, respectively. The TA impact category was primarily influenced by NH₃ emissions to the atmosphere from synthetic fertilizers and manure application, which generated an average of 67% and 27% of that contamination (kg SO₂ eq), respectively. Emissions of zinc (Zn) and copper (Cu) to water, the main drivers of FWE, were mainly derived from the manufacture of nitrogen synthetic fertilizers (74%).

In short, as can be seen in Figure 3, synthetic fertilizers were the main contributor to the selected impact categories (up to 87% in FWE) for the crops evaluated. More specifically, nitrogen synthetic fertilizer production had high impacts in all categories, with an average influence of 32% in all the crops analyzed.

Upon disaggregating the impacts per crop (Figure A2, Appendix B), wheat production stands out as the most impactful crop per ha FU, having the highest values in the CF, ME, and FWE impact categories, due to the higher amount of synthetic nitrogen fertilizers consumed compared to the rest (Table 2). The results per t FU show that alfalfa and corn silage had the lowest impacts, due to the high amount of mass produced (Table 2). Overall, oat, sunflower, and rapeseed were the crops with the highest impacts per t compared to the other crops, due to their low productivity (Figure A2, Table 2).

3.2.2. Animal Housing, Feed Management, and Manure Management

The most impactful subsystem and stage for each impact category are shown in

Table 4. Most impactful subsystem and stage, as well as the associated emissions among animal housing, feed management, and manure management (sum of the impacts = 100%).

Impact	Subsystem	Stage	Emissions	Compartment
CF	Animal housing (59.7%)	Enteric fermentation (56.3%)	CH4	Air
FE	Animal housing (56.4%)	Energy (electricity, 56.28%)	PO43−	Water
ME	Feed management (98.3%)	Soybean meal production (98.2%)	NO3−	Water
TA	Feed management (61.7%)	Soybean meal production (45.2%)	SO2, NH3	Air
FWE	Animal housing (48.5%)	Energy (electricity, 48.1%)	Cu	Water

, as well as the associated emissions for the three subsystems (animal housing, feed management, and manure management). The emitted CH₄ from fermentation in the animal housing was the main contributor (56.3%) to the CF impact category. The electricity consumed in the animal housing had the most major contributions in the FE (56.28%) and FWE (48.1%) categories, due to the high emissions to water of PO₄³⁻ and Cu, respectively. Feed management includes the production of exogenous feed and the energy used in the management of all feed. Exogenous feed had the highest weight in ME (98.3%) and TA (61.7%), primarily caused by the emissions derived from imported soybean meal (Table 4). None of the manure management stages stood out as being the mostly pollutant in a certain category, but it should be noted that the emissions of N₂O and CH₄ were the main sources of impacts caused by this subsystem.

4. Discussion

4.1. Comparison with Specialized Systems and Previous Studies

The impacts of the MFS studied have been compared with previous studies and against results of specialized systems to determine the extent to which the coupling of crop and livestock production provides a footprint improvement compared to specialization. Only the products sold by the farm (livestock products and cash crops) were compared with their counterparts in specialized systems.

• Livestock products

The CF is a popular indicator, which is often calculated to evaluate both crops and dairy cattle farms using the LCA methodology. The CF for milk production (kg CO₂ eq. kg FPCM⁻¹) of the MFS studied was 1.19 with no allocation and 1.17 after economic allocation (Table 3a). Overall, this value is below the mean CF value of the most widely adopted intensive milk production system in developed countries (1.29 kg CO₂ eq.), as well as the CF of extensive systems (6.6 kg CO₂ eq.) mainly used in developing countries [43,44]. Our results were slightly inferior than the average of 1.2 kg CO₂ eq. kg FPCM⁻¹ estimated by Mu et al. [45] for 55 specialized dairy farms from 6 European countries (Belgium, France, Germany, Ireland, Luxembourg, and the Netherlands). The CF obtained was also 27% lower than that found in the MFS in India by Paramesh et al., 2019 [20], due to the low productivity of milk in the studied system, as well as the higher dependency on the imported feed.

Relatively higher values of CF were found by Wang et al. [46], ranging from 1.31 to 2.08 kg CO₂ eq. kg FPCM⁻¹ with allocation when evaluating eight dairy farms in China for similar conditions, i.e., confined animals and having some similar feeds (e.g., soybean, maize, and wheat). In the cited study, high impacts were mainly attributed to low productivity (averaging around 18.5 kg/head/day), enteric emissions (54–60%), and off-farm feed production (21–30%) based entirely on synthetic fertilizers. The same work highlighted the significant increase in GHG emissions caused by the imported soybean meal, as also shown in this study, and the need to apply manure to crop production instead of exclusively using synthetic fertilizers. Previous MFS studies, which used data from statistical databases [4,18,19] and experimental sites [20,21] also revealed high impacts caused by synthetic fertilizers. Uddin et al. [44] also found higher CF values for four dietary scenarios, for which the minimum value was 1.31 and the maximum value was 1.56, with a mean value of 1.42 kg CO₂ eq. In that study, the farm purchased five feed ingredients (exogenous feed), whereas our studied system was based mainly on locally produced feed (85%). Biagetti et al. [47] reported that the average CF of three specialized dairy cattle farms in Italy was 1.77 kg CO₂ eq. kg FPCM⁻¹, considering that 60% of feed was sourced locally. Therefore, the better CF performance of the MFS studied was probably linked to the higher feed self-sufficiency (85%) and the lower reliance on synthetic fertilizers, by using manure for crop production.

However, it is important to note that enteric CH₄ is the major contributor to the CF (0.54 kg CO₂ eq in this study) and that it is greatly influenced by dietary composition [44]. The high CF from enteric fermentation has been widely highlighted in previous LCA studies of milk and beef production [20,21,46,47,48]. The CF could be further optimized with other animal management conditions, such as pasture-based systems and using better feed quality [48,49]. The latter could also be improved by the better dosing and management of fertilizers and cultivation practices, considering crop and soil requirements [50].

When looking at the specific impacts of milk and meat, the allocation factor for milk in the case studied was 99%, which is in line with the range of 90–100% indicated for the impacts of milk rather than meat [26,50]. This rate coincides with previous studies [28,50], which reported around 98–95% of environmental load for milk using economic allocation, respectively. Carvalho et al. [50] highlighted that the high rate attributed to milk for a semi-intensive and specialized system in Brazil may result in higher productive efficiency. The meat CF of the cited works [28,50] was higher than the MFS studied (3 × 10⁻² and 8.2 × 10⁻² vs. 1.19 × 10⁻² kg CO₂ eq. kg FPCM⁻¹).

Regarding eutrophication (FE and ME) and TA impacts, the MFS studied showed values per kg of FPCM of 2.52 × 10⁻⁴ kg P eq, 2.67 × 10⁻⁴ kg N eq, and 4.09 × 10⁻³ kg SO₂ eq, respectively. These values were much lower than the values calculated by Mu et al. [45] for the 55 specialized farms in 6 European countries, with an average per kg FPCM of 1.10 × 10⁻³ kg P eq, 8.10 × 10⁻³ kg N eq, and 2.61 × 10⁻² kg SO₂ eq, respectively. Such results are likely mainly due to the lower dependency on exogenous inputs such as feed (as in the case of CF), but also to the less intensive use of synthetic fertilizers on the MFS studied, which emphasizes the importance of reconnecting crop and livestock local systems and promoting circularity.

• Sold crops

Life cycle impacts per mass produced of the crops sold by the MFS studied (wheat, barley, and corn kernels) were compared with previous studies of specialized systems with comparable rainfed production methods.

The CF per t for the wheat cultivation in the studied case was 3.71 × 10² kg CO₂ eq. This is below previously reported values of CF ranging from 3.80 × 10² to 8.93 × 10² kg CO₂ eq t⁻¹ [51,52,53,54]. The better environmental performance of wheat grown under the MFS studied is also highlighted in the other selected impact categories (FE, ME, TA, and FWE) compared to those obtained for specialized systems in other studies [52,54]. In fact, the MFS studied caused 5.30 kg SO₂ eq t⁻¹ in the acidification and 2.11 × 10⁻¹ kg P eq t⁻¹ in the eutrophication (FE), whereas the values for rainfed systems in previous comparable studies [52,54] were 11.86 and 40.31 kg SO₂ eq t⁻¹, and 1.06 and 2.9 × 10⁻¹ kg P eq t⁻¹, respectively. This result was mainly related to a lower reliance on nutrients from synthetic fertilizers, since local manure provided 15%, 20%, and 50% of N, P, and K needs, respectively, as well as the registered higher productivity. However, the environmental performance of wheat cultivation could be further improved by using better nitrogen fertilizer management, as shown by Todorović et al. [41]. These authors showed that optimizing water and fertilizer inputs (i.e., deficit irrigation with low nitrogen) together with the adoption of precision agriculture had a greener footprint than highly intensive cropping strategies.

As regards previous LCA studies for rainfed barley and grain maize, 53 farms were evaluated in Iran [55] and 18 ha of barley were evaluated in Norway [56]. Both studies had higher CF per kg of barley grain than in the case study, with 1.31 and 7.94 × 10⁻¹ kg CO₂ eq., respectively (vs. 3.53 × 10⁻¹ kg CO₂ eq. kg⁻¹ for the MFS studied). The same result was observed for the other selected categories (eutrophication and acidification), except for freshwater ecotoxicity compared to the barley in Norway [56]. As in the case of wheat, the environmental outperformance of the MFS studied is likely due to the lower exogenous inputs.

4.2. Interpretation and Extrapolation of Results

Overall, the MFS studied had a somewhat better environmental performance compared with other specialized systems in different areas across the world. However, our results evidenced that to further improve the footprint of the MFS, some key impact factors could be improved, such as enteric methane emissions and exogenous nutrients. The imported feeds and synthetic fertilizers were mainly influenced by feed composition and efficiency and management conditions [49,57]. For example, herd management based on grazing and offering concentrates to dairy cows [49], particularly concentrates containing linseed oil [57], could reduce CH₄ emissions (up to 18%) per unit of milk compared with those containing stearic acid or soy oil.

This study contributes to the development of a life cycle methodological framework for evaluating complex circular systems. An important aspect to be highlighted when interpreting the results is that, for fair comparisons, they should be compared to counterpart systems with equivalent characteristics (crops, animals, and climate conditions) and using similar assumptions (emissions, boundaries, etc.). Therefore, data should be interpreted cautiously, since LCA models present great variability and may have different system boundaries, impact assessment methods, assumptions and FUs. With respect to previous MFS studies, comparisons are limited by the different FUs used, depending on the objective of the study (e.g., megajoules of digestible energy for humans [18]), and the animals and products analyzed (e.g., pig [19]), as well as the ways in which the impact results have been presented (e.g., in percentages [4]). When contrasting results with LCA studies on livestock production, it is to be noted that they often assume that exogenous feeds are produced on the target farm, which fails to reflect the level of circularity and hence may compromise comparability. Therefore, studies not clearly identifying and modeling feed imported by the farm as an exogenous input would not be suitable as benchmarks.

More real case studies at both a local and regional level are needed, in order to understand in a broader context, the impact of reconnecting livestock and crop production. However, the marked reduction in these systems, especially in Romania, may hinder the collection of detailed data for input-intensive methodologies such as LCA. Nevertheless, the present case example shows how feed self-sufficiency and manure recycling in a connected livestock-crops system can mitigate multiple environmental impacts including carbon footprint, eutrophication, and acidification and freshwater ecotoxicity. To further attract policy makers’ attention, future work should include social and economic aspects, as there may be competing interests of which the interdependencies should be addressed.

5. Conclusions

This study assessed the LCA impacts of a coupled livestock–crop commercial farm in Romania. The assessment included eight feed crops, three cash crops, and two livestock products (milk and meat). This is a relevant case study, particularly considering that Romania is one of the European countries where the number of mixed systems has significantly declined in the last decade. The sustainability of mixed farms is a key topic of debate and discussion in Europe, which is a world leader in livestock and dairy cattle production.

Overall, the present study shows that the reconnection of livestock and crop production has considerable potential to improve the agricultural footprint by reducing the over-reliance on environmentally burdensome stages such as synthetic fertilizers and exogenous feed. In particular, our results show that (i) although the mixed farm had a lower footprint that comparable specialized systems, the impacts were still relatively high due to the need to import synthetic fertilizers and exogenous feed and, therefore, further self-sufficiency should be promoted and (ii) methane emissions from enteric fermentation remain the major carbon footprint burden for livestock production and, hence, reducing this impact remains a priority to ensure better sustainability. This issue could be addressed through sustainable management, including grazing of animals, higher quality feed, and better use and management of farm resources.

Finally, our results could prove useful not only for the Romanian livestock sector but can also provide a benchmark for other regions in Europe, especially with regard to confined animals. Further studies should address mixed farming systems under other conditions, such as pasture-based systems and systems with crop rotations, which have more interactions and synergies between farm components and probably better environmental outcomes. Economic and social dimensions should also be considered to provide a complete sustainability assessment.