Cover crops support the climate change mitigation potential of agroecosystems

RESEARCH ARTICLE Cover crops support the climate change mitigation potential of agroecosystems Jonas Scho¨ n1, Norman Gentsch2, Peter BreunigID1* 1 Weihenstephan-Triesdorf University of Applied Sciences, Triesdorf, Germany, 2 Institute of Earth System Science, Section Soil Science, Leibniz Universita¨t Hannover, Hannover, Germany * peter.breunig@hswt.de Abstract Cover crops have the potential to mitigate climate change by reducing negative impacts of agriculture on ecosystems. This study is first to quantify the net climate change mitigation impact of cover crops including land-use effects. A systematic literature and data review was conducted to identify major drivers for climate benefits and costs of cover crops in maize (Zea maize L.) production systems. The results indicate that cover crops lead to a net climate change mitigation impact (NCCMI) of 3.30 Mg CO2e ha-1 a-1. We created four sce- narios with different impact weights of the drivers and all of them showing a positive NCCMI. Carbon land benefit, the carbon opportunity costs based on maize yield gains following cover crops, is the major contributor to the NCCMI (34.5% of all benefits). Carbon seques- tration is the second largest contributor (33.8%). The climate costs of cover crops are mainly dominated by emissions from their seed production and foregone benefits due to land use for cover crops seeds. However, these two costs account for only 15.8% of the benefits. Extrapolating these results, planting cover crops before all maize acreage in the EU results

in a climate change mitigation of 49.80 million Mg CO2e a-1, which is equivalent to 13.0% of the EU’s agricultural emissions. This study highlights the importance of incorporating cover crops into sustainable cropping systems to minimize the agricultural impact to climate change. 1. Introduction Mitigating climate change through measures in agriculture and food systems is essential for achieving the 1.50˚ and 2.00˚C climate change targets of the Paris Agreement [1]. Worldwide, agriculture and food production systems are responsible for about one-third of global anthro- pogenic greenhouse gas (GHG) emissions, of which 40.0% are caused by agricultural produc- tion and 32.0% by land use and land use change [2]. However, agricultural systems also offer opportunities for climate change mitigation. Agroecosystems have significant potential to reduce global net emissions [3] and could even act as a net carbon sink [4]. To achieve these emission reduction and carbon sequestration opportunities, a transformation of agricultural systems is required. An essential component of climate-smart agricultural systems is planting cover crops. Cover crops are plants or plant mixtures that are not harvested for revenue, but PLOS ONE PLOS ONE | https://doi.org/10.1371/journal.pone.0302139 May 8, 2024 1 / 20 a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 OPEN ACCESS Citation: Scho¨n J, Gentsch N, Breunig P (2024) Cover crops support the climate change mitigation potential of agroecosystems. PLoS ONE 19(5): e0302139. https://doi.org/10.1371/journal. pone.0302139

Editor: Abhay Omprakash Shirale, IISS: Indian Institute of Soil Science, INDIA Received: November 29, 2023 Accepted: March 28, 2024 Published: May 8, 2024 Peer Review History: PLOS recognizes the benefits of transparency in the peer review process; therefore, we enable the publication of all of the content of peer review and author responses alongside final, published articles. The editorial history of this article is available here: https://doi.org/10.1371/journal.pone.0302139 Copyright: © 2024 Scho¨n et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Data Availability Statement: All relevant data are within the manuscript and its Supporting Information files. Funding: The research was funded by the German Federal Ministry of Education and Research within

are grown for the purpose of soil nutrient management, organic matter input, soil protection, and soil health improvement [5]. Cover crops are associated with multiple ecosystem services, such as closer nutrient cycling [5,6], activation of soil biology, biodiversity improvement [7], soil water management, and restoration of soil structures [8]. In addition, the mentioned effects of cover crops could mitigate climate change by soil carbon sequestration [9], biomass carbon storage, or reduced fertilizer losses to aquatic ecosystems. Recent studies show that these effects can lead to agronomic benefits that translate into higher yields and reduced agri- cultural input [10]. Despite these benefits, there exists a knowledge gap. The research problem addressed in this article is that there are no complete climate impact assessments of the cultivation of cover crops that also fully include land use effects. Cover crops reduces land use requirements by increasing yields of the subsequent crop. As land use change is still the largest carbon emis- sions source in the agri-food-system, yield increasing measures can be considered as a climate change mitigation option. According to Searchinger et al. [11], these carbon land benefits can be calculated as a “carbon opportunity cost,” which is the foregone carbon storage potential from natural vegetation to produce a certain agricultural product in kg CO2e kg-1. When yields rise on one piece of land, this carbon storage potential can be maintained or restored on other

locations since land use change is prevented or reverted. Kovak et al. [12] used this approach to quantify the climate change mitigation potential of a yield increase based on genetic modi- fied crops. This is the first study to include carbon opportunity cost in the evaluation of the cli- mate impact of cover cropping. As an additional climate benefit, cover crops remove carbon from the atmosphere during their growth while their residues and rhizodeposits are trans- formed to stable soil organic matter by retaining soil moisture. Further climate benefits of cover crops can be attributed to the supply of nutrients to the subsequent crop, which allows for fertilizer savings [13], reduced N2O emissions due to less nitrogen leaching, reduced solar radiation due to plant growth [14], and reduced soil erosion [15]. We define all factors above as “climate benefits”. While cultivating cover crops can accrue these climate benefits, it can also generate GHG emissions that we refer as “climate costs”. The cultivation of cover crops can increase N2O emissions in periods of anaerobic microbial conversion of cover crop decomposition products [16]. Emissions are also generated in the production of cover crop seeds through a number of avenues, such as the land required for seed production and the production process of the seeds. In addition, produced seeds need to be packed and delivered, which also generates emis- sions[17]. Lastly, tilling, seeding, and termination of cover crops requires the use of GHG

emitting machinery and resources [18]. Both climate benefits and climate costs of cover crops need to be carefully considered to calculate the NCCMI. This article addresses the research problem by developing a comprehensive framework to quantify the NCCMI of cover crops including land use effects based on a systematic literature review. Using selected keywords, a systematic search was conducted in “Web of Science” and “Google Scholar” and resulted in 1269 publications. We define a list of relevant climate bene- fits and climate costs from which the NCCMI of cover crops is derived. We apply the frame- work to cover crops for maize in the European Union (EU-27, which is all EU Member States excluding the United Kingdom) to quantify its total NCCMI. 2. Methods The following analyses are based on a typical Central European crop rotation, namely winter cover crops that are incorporated into the soil and maize as a cash crop following the cover crops in the spring. To quantify the impact of cover crop cultivation on the climate mitigation PLOS ONE Cover crops support the climate change mitigation potential of agroecosystems PLOS ONE | https://doi.org/10.1371/journal.pone.0302139 May 8, 2024 2 / 20 the Project "CATCHY", project number: 031B1060C. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing interests: The authors have declared that no competing interests exist.

potential, a systematic literature review was conducted. The research platforms "Web of Sci- ence" and "Google Scholar" were used to explore the literature. Based on the initial research, we defined five factors that support climate change mitigation (“climate benefits”) and five fac- tors leading to additional GHG emissions (“climate costs”) due to cover cropping (Table 1). The individual research terms of the individual impact factors are described below in detail. To calculate the NCCMI of cover crops, the values of all impact factors were converted to the same unit, namely megagram of CO2 equivalents per hectare and year (Mg CO2e ha-1 a-1). From the values in the literature, a weighted average value that takes into account the number of measurements in the specific study was calculated. Only data from the literature that indi- cated the number of measurement points were considered in the calculation of the NCCMI. Thus, the climate benefits and climate costs could be added together in each case before sub- tracting the costs from the benefits. This resulted in a value for the NCCMI per hectare of cover crops given in Mg CO2e ha-1 a-1. The 95% confidence interval was calculated from all study means and presents a range of estimates. To extrapolate the results to the EU-27 maize area the respective impact factors and the climate balances were subsequently multiplied with the maize acreage in the EU-27. The result was then divided by the total emissions from agri- culture in the EU-27. This calculation resulted in the share of the NCCMI of cover crops before

maize relative to all EU-27 agricultural emissions. The average maize acreage in 22 of the EU- 27 countries was calculated as an average based on the years 2011 to 2022 (five of the EU-27 countries do not grow relevant quantities of maize and were thus not considered.) The following search terms were used in Web of Science: Cover crop & albedo, Cover crop & C sequestration, Cover crop & Yield corn, Cover crop & Yield maize, Cover crop & life cycle assessment, Cover crop & N-fixation, Cover crop & N- leaching, Fertilizer & Greenhouse gas emissions, Greenhouse gas emissions & fertilizer pro- duction, Cover crop & N2O, Catch crop & N2O. To enable a phrase search, all search terms were used with quotation marks to obtain search results exactly in this word order. A total of 883 publications were extracted and checked for their suitability. Criteria for the required data were an absolute value from the respective study and the number of measure- ments. Only 44 articles met these criteria and were finally included in the review. In addition, we used 31 publications from Google Scholar and data from 26 other Internet sources that were selected based on the above criteria. In selecting sources, cover crops were considered both consisting of individual components and mixtures, overwintering as well as freezing, and legumes as well as non-legumes. All data sources are presented in the supplemental material (S1 File) and sorted based on the subsections below. 2.1. Carbon land benefit based on yield gain

The relative yield gain of maize following cover crops is based on literature research. The abso- lute yield gain was calculated by multiplying the relative yield gain with the average maize yield in the EU-27. For this purpose, the harvest volume of silage and grain maize in Europe Table 1. Calculation for mean land use for cover crops’ seed production. Species Seed yield in kg ha-1 Seeding rate in kg ha-1 Seed land use in ha seed ha-1 for cover crops Phacelia 350.00 10.50 0.03 Mustard 1,738.00 20.00 0.01 Clover 732.00 29.17 0.04 Oats 4,869.50 87.50 0.02 Mixture 0.03 https://doi.org/10.1371/journal.pone.0302139.t001 PLOS ONE Cover crops support the climate change mitigation potential of agroecosystems PLOS ONE | https://doi.org/10.1371/journal.pone.0302139 May 8, 2024 3 / 20

was divided by the cultivated area. Since carbon opportunity costs factors published in Search- inger et al. [11] refer only to grain maize, the silage maize (SM) yields were transformed to grain maize (GM). It was assumed that 54% of the dry matter of silage maize is grain [19]. Thus, the fresh mass of silage corn was multiplied by a factor of 0.35 to calculate the dry matter and then multiplied by 0.54 [19] to generate the grain yield of SM (see the supplemental mate- rial for more detail). To obtain the grain yield based on the standard of 14% moisture, the result was divided by 0.86. The following formula shows the complete calculation to transfer SM to GM: GM ¼ SM � 0:35 � 0:54 0:86 : To quantify the effect of yield increases on climate change mitigation, we use the carbon land benefit approach of Searchinger et al. [11]. This approach quantifies the benefit of increas- ing yields on one piece of land through the carbon storage potential of native vegetation on other land that is not required to produce a certain quantity of product because of higher pro- duction on existing agricultural land. According to Searchinger et al. [11], the carbon land benefit carbon opportunity costs (COCs) is defined as follows: COCs ¼ Y � COC; where, COCs is the carbon land benefit, Y is the grain and biomass yield of an agricultural product, and COC is the carbon opportunity cost factor for the agricultural product. COC quantifies the forgone carbon storage potential of the native vegetation of land, which is glob-

ally used to produce an agricultural commodity and given in kg CO2e per kg of product. Searchinger et al. [11] use two methods to quantify the carbon opportunity cost. Here, we use the values of the so-called “loss-method.” To quantify the carbon land benefit caused by yield gains in maize through cover crops, the absolute yield increase of maize in kg ha-1 is multiplied by 2.1 kg CO2e kg-1, which is the carbon opportunity cost factor of grain maize published in Searchinger et al. [11]. 2.2. Carbon sequestration Carbon sequestration through cover crops is usually provided in units of soil organic carbon. This value was multiplied by a factor of 3.667 (i.e., the ratio of molar masses of CO2 and C) to obtain the value of CO2 stored per ha land. 2.3. Nitrogen fertilizer savings For nutrient savings, we considered only the main nutrient, namely nitrogen since its produc- tion is highly energy intensive. To quantify fertilizer savings, we only considered emissions of mineral fertilizers and not that of organic fertilizers. The overall nitrogen fertilizer savings were quantified as the sum of the mean literature values on nitrogen scavenging and the mean literature values of nitrogen fixation from cover crops. 2.4. Reduced indirect N2O emissions due to less leaching Reducing nitrogen leaching due to cover crops leads to lower indirect N2O emissions since less nitrogen is deposited to rivers and other water bodies. To quantify this effect, we use the approach suggested in the 2006 IPCC Guidelines for National Greenhouse Gas Inventories:

The amount of nitrogen leaching reduction is multiplied by the emission factor EF5 of 0.0075 [20] to obtain the reduction in indirect N2O emissions per hectare. This value is then con- verted from kg N2O-N ha-1 to kg CO2 -eq ha-1 using the conversion factor 273 and divided by PLOS ONE Cover crops support the climate change mitigation potential of agroecosystems PLOS ONE | https://doi.org/10.1371/journal.pone.0302139 May 8, 2024 4 / 20

1000 to obtain the unit Mg CO2 -eq ha-1 [21]. Reduced N2O N ¼ EF5∗Reduced NLeaching ∗273 1000 2.5 Albedo change The reduced adsorption and thus increased reflection of solar radiation by cover crops’ plant cover compared to fallow land leads to a climate change mitigation effect. We found, however, only two studies evaluating the impact of albedo change from cover crops (Kaye and Quemada [22] and another by Carrer et al. [14]). 2.6. Nitrous oxide emissions In the case of nitrous oxide emissions, the cumulative emissions that occur during the period of fallow or the period from sowing to incorporation of cover crops are considered. The values with the unit kg N2O-N ha-1 were multiplied by a factor of 273 and 1000 to obtain the unit in Mg CO2e ha-1. 2.7. Foregone benefits due to cover crop seed land use The production of seeds for cover crops requires additional land that is not available for pro- duction of agricultural products. Land for these seeds cannot create the same carbon benefits [11] as alternative uses of this land. Therefore, this forgone carbon benefit needs to be added to the climate costs of cover crops. For this purpose, the yields of the four most common seed components of a cover crop mixture were determined in order to calculate the average land requirement for seed production of one hectare of cover crops’ seed mixture based on typical seeding rates. The following four species were selected to determine the land use for cover crops’ seed production: phacelia (Phacelia tanecetifolia Benth.), mustard (Sinapis alba L.), clo-

ver (Trifolium alexandrinum L.), and oats (Avena strigose Schreb.). Each of these cover crops has a 25% share in the final mixture of cover crops. A literature search on seed yields and aver- age seeding rates for these four species was conducted. The seed land use (SLU) for a specific cover crop was calculated per ha based on the following formula: SLU ¼ 1 n Xn i¼1 Ycc SRcc � � i; where Ycc is the yield in kg ha-1 of a specific cover crop and SRcc is the seeding ratio in kg ha-1. Table 2 shows the mean input values derived from the literature for the land use require- ments for cover crop seeds. The forgone carbon benefit of the land used for the production of cover crop seeds is deter- mined as follows: We assume that seeds for cover crops are produced in the European Union on land that would otherwise be used for wheat production since wheat is the largest crop in the EU by acreage. The carbon benefit (CB) of wheat is defined by Searchinger et al. [11]. Here, we assume that the carbon stock of land is not changing between wheat and seed pro- duction for cover crops and that there are no carbon benefits from bioenergy. The equation for CB is as follows: CB ¼ ðCOC þ PEMavg PEMEU Þ � Y where COC is the global carbon opportunity cost factor of wheat, PEMavg is the global average PLOS ONE Cover crops support the climate change mitigation potential of agroecosystems PLOS ONE | https://doi.org/10.1371/journal.pone.0302139 May 8, 2024 5 / 20

Table 2. Calculation of the net climate change mitigation impact of cover crops during an annual cropping cycle. EU-27 average per hectare (ha) EU-27 total Unit Mean Low (-95% confidence level) High (+95% confidence level) Unit Mean Low (-95% confidence level) High (+95% confidence level) Area 1,000 ha 15,092.14 14,785.60 15,398.67 Average yield Mg ha-1 7.88 7.46 8.31 Relative yield gain % 8.8% 2.9% 14.8% Absolute yield gain Mg ha-1 0.70 0.21 1.23 Carbon opportunity cost maize kg CO2e kg-1 2.10 2.10 2.10 Carbon land benefit based on maize yield gain Mg CO2e ha-1 1.46 0.45 2.59 1,000 Mg CO2e 22,086.83 6,631.87 39,805.69 Carbon sequestration Mg CO2 ha-1 1.43 0.86 2.01 1,000 Mg CO2 21,632.51 12,701.51 30,915.61 Nitrogen fixation kg N ha-1 52.49 32.06 72.92 Reduced nitrogen leaching kg N ha-1 38.82 19.13 58.51 Total nitrogen fertilizer savings kg N ha-1 91.32 51.19 131.44 Nitrogen fertilizer emissions kg CO2e kg N-1 11.23 9.02 13.43 Carbon benefit based on nitrogen fertilizer savings Mg CO2e ha-1 1.03 0.46 1.77 1,000 Mg CO2e 15,474.08 6,830.39 27,187.04 Reduced N2O emissions due to less leaching Mg CO2e ha-1 0.10 -0.15 0.35 1,000 Mg CO2e 1,495.60 -2,218.43 5,362.37 Albedo change Mg CO2e ha-1 0.20 -0.37 0.78 1,000 Mg CO2e 3,087.10 -5,514.25 12,042.49 Total climate benefit Mg CO2e ha-1 4.22 1.25 7.49 1,000 Mg CO2e 63,776.11 18,431.09 115,313.18 N2O emissions Mg CO2e ha-1 0.04 -0.02 0.09 1,000 Mg CO2e 555.77 -306.57 1,453.41 Foregone benefits due to cover crop seed land use Mg CO2e ha-1

0.28 0.27 0.29 1,000 Mg CO2e 4,180.56 3,970.76 4,395.55 Production emissions cover crop seed Mg CO2e ha-1 0.39 0.08 0.70 1,000 Mg CO2e 5,920.55 1,194.38 10,837.69 Processing, packaging, and transport of cover crop seed Mg CO2e ha-1 0.08 0.08 0.08 1,000 Mg CO2e 1,182.71 1,158.69 1,206.74 Additional machinery operations l Diesel ha-1 45.29 32.44 58.13 Diesel emissions kg CO2 l-1 3.13 3.13 3.13 Additional machinery operations Mg CO2 ha-1 0.14 0.10 0.18 1,000 Mg CO2e 2,140.93 1,502.51 2,804.02 Total climate cost Mg CO2e ha-1 0.93 0.51 1.34 1,000 Mg CO2e 13,980.53 7,519.77 20,697.40 (Continued) PLOS ONE Cover crops support the climate change mitigation potential of agroecosystems PLOS ONE | https://doi.org/10.1371/journal.pone.0302139 May 8, 2024 6 / 20

production emissions for wheat, PEMEU is the European Union’s production emissions for wheat, and Y is the average wheat yield in the European Union (based on data from 2011 to 2022 for the EU-27 [23]). COC, PEMavg, and PEMEU are from Searchinger et al. [11]. For PEMEU, we used the production emissions for wheat in Sweden, since no European average values were available. The forgone carbon benefits due to seed land use can be calculated by multiplying SLU and CB. 2.8. Production emissions for cover crop seed production, dispatch and additional machinery costs Product emissions for cover crop seed production are not available in literature and therefore derived from maize, rapeseed, and spring wheat seed production. This is a conservative assumption since the nitrogen input and machinery operations as key drivers of emissions are usually higher in these crops than in cover crops. Emissions data is collected from the literature in kg CO2e kg-1 of seed and is multiplied with the seeding rate of cover crops. This results in an emissions value per hectare. Emissions for processing, packaging, and transporting cover crop seeds are also not avail- able and derived from maize seed production emissions and are expressed in kg CO2e kg-1 of seed. This is again a conservative assumption since maize seed has to fulfill much higher qual- ity standards than cover crop seed. The emissions of additional machinery operations for tillage, seeding, and mechanical ter- mination of cover crops were included only in terms of additional diesel fuel usage. Average

diesel fuel consumption of all additional operations was calculated and multiplied by the aver- age GHG emissions of diesel fuel. 3. Results The impact of all analyzed factors on the NCCMI of cover crops is summarized in Table 3. The results refer to a typical European crop rotation using a winter cover crop and maize Table 2. (Continued) EU-27 average per hectare (ha) EU-27 total Unit Mean Low (-95% confidence level) High (+95% confidence level) Unit Mean Low (-95% confidence level) High (+95% confidence level) Net climate change mitigation impact of cover crops Mg CO2e ha-1 3.30 0.74 6.14 1,000 Mg CO2e 49,795.58 10,911.32 94,615.78 Share of EU-27 agriculture emissions in 2020 13.0% 2.9% 24.7% Note: Calculation, formulas and number of studies are provided as supplementary material (S1 File). ha is hectares. EU-27 is the 27 Member States of the European Union (all Member States excluding the United Kingdom). https://doi.org/10.1371/journal.pone.0302139.t002 Table 3. Impact factors of cover crops on climate change mitigation analyzed in this study, source: Own table. Climate Benefits Climate Costs • Carbon land benefit based on yield gain • N2O emissions • Carbon sequestration • Foregone benefits due to cover crop seed land use • Nitrogen fertilizer savings • Production emissions for cover crop seed production • Reduced indirect N2O emissions due to less leaching • Processing, packaging, and transport of cover crop seed • Albedo change • Additional machinery operations https://doi.org/10.1371/journal.pone.0302139.t003 PLOS ONE

Cover crops support the climate change mitigation potential of agroecosystems PLOS ONE | https://doi.org/10.1371/journal.pone.0302139 May 8, 2024 7 / 20

erature research revealed two meta-analyses with 139 and 1195 studies included. In addition, 26 studies that were not included in the above-mentioned meta-analyses were also included in our calculations. Up to 12.0% of carbon from cover crop biomass can be sequestered as soil organic carbon [13]. Similar to the crop yield effects shown above, the potential of soil organic Fig 1. Forest plot of included literature on yield effects. https://doi.org/10.1371/journal.pone.0302139.g001 PLOS ONE Cover crops support the climate change mitigation potential of agroecosystems PLOS ONE | https://doi.org/10.1371/journal.pone.0302139 May 8, 2024 8 / 20

carbon stock increases depending on the type of cover crop. For example, a soil organic carbon increase of 0.26 Mg CO2e ha-1 a-1 has been found for a winter vetch cover crop [13]. A much larger increase in soil carbon of 5.12 Mg CO2e ha-1 a-1 was found by Abdalla et al. [25] for non-legume cover crops. The weighted mean of all studies investigated shows an average sequestration of 1.43 Mg CO2e ha-1 a-1 (95% CI, 0.86, 2.01). Based on the included studies, grass cover crops lead to a 2.3-times higher soil organic carbon sequestration than legume cover crops. Cover crop mixtures of grasses, legumes, and various other species are in the mid- dle of the field in sequestration performance [26]. Fig 2 shows the included carbon sequestration results, confidence intervals and the pooled effect. If more than one result is shown from one study in Fig 2 this literature showed results from multiple experiments. 3.3. Nitrogen fixation Legume cover crops are able to fix atmospheric nitrogen and transfer it to the rhizosphere or utilize it for their own biomass. The decay of nitrogen-rich legume litter or rhizosphere prod- ucts contribute to soil nitrogen fertilization and plant nutrition of the crop following the cover crop [13]. The amount of nitrogen fixed depends on the legume species and/or the companion Fig 2. Forest plot of included literature on carbon sequestration effects. https://doi.org/10.1371/journal.pone.0302139.g002 PLOS ONE Cover crops support the climate change mitigation potential of agroecosystems

PLOS ONE | https://doi.org/10.1371/journal.pone.0302139 May 8, 2024 9 / 20

crop species in the mixture. For example, red clover fixes 29.00 kg N ha-1 as a single species cover crop, but 52.00 kg N ha-1 in a mixture with ryegrass [13]. The same was observed with winter vetch, which fixes 85.00 kg N ha-1 as single species and 113.00 kg N ha-1 in a mixture with winter rye [27]. A much wider range from 50.00 kg N ha-1 to 300.00 kg N ha-1 was found by Kaye and Quemada [22]. Wittwer et al. [18] showed that legumes as a single species cover crop can increase nitrogen uptake of maize by 32.00 kg N ha-1 and as part of cover crop mix- tures by 28.00 kg N ha-1. The summarized literature resulted in a weighted mean of 52.49 kg N ha-1 (95% CI, 32.06, 72.92) that can be contributed by legumes from nitrogen fixation. 3.4. Reduced nitrogen leaching Cover crops prevent nutrient leaching over winter, especially of mineral nitrogen (Nmin) that otherwise would be leached from the soil profile. However, this leaching reduction strongly depends on the timing of when cover crop are planted and type of cover crop. Early sown oil radish reduces leaching in the fall and winter months [28]. Rye as an over-wintering cover crop can better protect nitrogen from leaching in the spring [28]. Non-legume cover crops and mixtures can protect up to 100.00 kg N ha-1 from leaching, while pure legumes can protect between 60.00 to 70.00 kg N ha-1 [6,22]. Abdalla et al. [25] show that legume cover crops pre- vent 18.00 kg N ha-1 from leaching, non-legume cover crops 38.00 kg N ha-1, and cover crop mixtures 19.00 kg N ha-1. A study by McCracken et al. [29] resulted in an average leaching

reduction of 35.80 kg N ha-1. The amount of scavenged nitrogen depends heavily on water availability and soil type. Early sowing and sufficient precipitation enable cover crops to develop faster, establish more biomass, and absorb more nitrogen in the fall. The tillage system also affects the amount of scavenged nitorgen. The faster mineralization in conventional tillage systems in autumn resulted in higher Nmin stocks and higher nitrogen uptake from cover crops than in no-till systems [30]. Literature research resulted in a weighted mean value of 38.82 kg N ha-1 (95% CI, 19.13, 58.51) of scavenged nitrogen available to the crop following the cover crop. 3.5. Fertilizer emissions Nitrogen fertilizer emissions are caused by different factors and vary between fertilizer types. About half of the CO2e emissions of nitrogen fertilizer is N2O emissions from soil after appli- cation and about one-third of all emissions occur during the production process of fertilizer itself. The remaining emissions are divided into N2O emissions from NH4+ and NO3- losses and CO2 from lime [31]. Total emissions per kg of nitrogen also vary between fertilizer types and the literature. For example, Hasler et al. [32] reported 8.69 kg CO2e kg-1 N for urea fertil- izer to 11.64 kg CO2e kg-1 N for CaH4N4O9 fertilizer. A much higher emission factor of 15.00 to 31.00 kg CO2e kg-1 N is used by Kahrl et al. [33]. The mean value of all studies considered was 11.23 kg CO2e kg-1 N (95% CI, 9.02, 13.43). Emissions from organic fertilizers were not considered.

3.6. Reduced indirect N2O emissions due to less leaching Leached nitrogen is deposited into aquatic ecosystems and represents a globally significant source of N2O through indirect emissions. These emissions can be calculated using the IPCC emission factor EF5, which incorporates three components: EF5g (groundwater and surface drainage), EF5r (rivers) and EF5e (estuaries) [34]. The IPCC emission factor EF5 amounts to 0.0075 and is the sum of EF5g, EF5r and EF5e with a value of 0.0025 each. Tian et al. [34] suggest a higher emission factor for EF5g of 0.0060, the other two factors are almost the same as those suggested by IPCC in 2006. Here, we use the PLOS ONE Cover crops support the climate change mitigation potential of agroecosystems PLOS ONE | https://doi.org/10.1371/journal.pone.0302139 May 8, 2024 10 / 20

mean of two results: One using the IPPC factor which results in 0.08 Mg CO2 -eq ha-1 and the other one using the factor of Tian et al. [34] which results in 0.12 Mg CO2 -eq ha-1. The mean of both values leads to an average climate benefit caused by reduced indirect N2O emissions due to less leaching of 0.10 Mg CO2 -eq ha-1 (95% CI, -0.15, 0.35). 3.7. Albedo change When soil is covered with cover crops, less solar radiation reaches the soil surface compared to fallow land. Thus, the albedo value of plants is higher than that of fallow soil. Due to this reduced net radiation, there is a cooling effect that impacts the global climate. However, the albedo value of fallow soil strongly depends on soil type and condition. In particular, the color and moisture content of the soil play important roles. Dark and moist soils have a lower albedo than light dry soils, so there is a wide variation in the albedo value of fallow land. Since there is often a high soil moisture level during the growth period of cover crops (in fall and winter) the albedo effect of covering the soil with cover crops is stronger than for the yearly average [14]. Unfortunately, research on albedo effects of cover crops is in its infancy and only two adequate publications were found. The albedo effect of cover crops ranges from -4.00 to 10.00 W m-2 was calculated by Kaye and Quemada [22]. Converted to CO2e, this corresponds to a climate benefit of 0.25 Mg CO2e ha-1 a-1. Carrer et al. [14] show that three months of cover cropping

in Europe results in a climate benefit of 0.16 Mg CO2e ha-1 a-1. Similarly, this study found that longer time periods with cover cropping would increase the mitigation potential by 27.0%. Overall, a mean climate benefit of 0.20 Mg CO2e ha-1 a-1 (95% CI, -0.37, 0.78) was identified in the considered literature for the albedo change of cover crops. 3.8. Nitrous oxide emissions Cover cropping can increase N2O emissions under certain weather conditions compared to fallow land depending on the fertilizer N rate, soil incorporation, and the period of measure- ment and rainfall [35].There are many factors that influence the level of N2O emissions which leads to high variability from this GHG emission source. A prerequisite for cover crops is high soil moisture and water saturation, or at least water saturation of the upper soil layers. Another factor is the C/N ratio of the cover crop, which influences how fast cover crop residues decom- pose. Cover crops with low C/N ratios emit more N2O than cover crops with high C/N ratios [36,37]. Another factor influencing N2O emissions is the treatment of cover crop residues. Incorporating residues into the soil leads to higher N2O emissions as compared to leaving cover crop residues decaying on the soil surface [36]. Air temperature also effects N2O emis- sions. As temperatures increase, N2O emissions from cover crops also increase. Overall, air temperature accounts for 22.0% of the variability in N2O emissions [38]. In a meta-analysis, Basche et al. [35] found that N2O emissions increased in 60.0% and decreased in 40.0% of the

publications when cover cropping was practiced. The overall conclusion of the meta-analysis is that cover cropping increases N2O emissions only to a very small extent and that the vari- ability is very large. For example, compared to a chemical fallow system, Olofsson and Ernfors [39] identified the highest value with 1.80 kg N2O-N ha-1 for oilseed radish, whereas Preza- Fontes et al. [40] identified the lowest emissions with -0.54 kg N2O-N ha-1 for a sorghum- sudan-grass-mixture. The result of our calculation based on a systematic literature search also shows only a very small increase in N2O emissions from cover cropping. The determined weighted mean value of emissions, as displayed in Table 2, is 0.13 kg N2O-N ha-1, which is equivalent to 0.04 Mg CO2e ha-1 (95% CI, -0.02, 0.09). PLOS ONE Cover crops support the climate change mitigation potential of agroecosystems PLOS ONE | https://doi.org/10.1371/journal.pone.0302139 May 8, 2024 11 / 20

3.9. Foregone benefits due to cover crop seed land use Land use for producing seeds for cover crops was calculated for the species phacelia, mustard, clover, and oat as representatives of the most important plant families used as cover crops. Their average was used for further calculations. The average land use requirement to produce one ha of cover crop seeds was calculated as 0.02 ha (95% CI, 0.00, 0.04). We assume that the land used for the production of cover crop seeds would otherwise be used for wheat production. Wheat production on land in the EU offers a carbon benefit that consists of: (1) the opportunity that the wheat output enables storing carbon elsewhere (yield × carbon opportunity cost); and (2) savings in global production emissions due to lower production emissions in the EU compared to the global average (here, the difference between Swedish and global wheat production emissions is based on Searchinger et al. [11]). Other fac- tors, like changes in soil organic carbon or bioenergy benefits are not considered here. The cli- mate benefit of wheat consisting of factors (1) and (2) is forgone due to land requirements for cover crops’ seed requirements and can be quantified accordingly to a mean value of 0.28 Mg CO2e ha-1 (95% CI, 0.27, 0.29). 3.10. Production emissions for cover crop seed production Since no specific literature sources for seed production emissions could be found for typical cover crop species, emissions from corn, spring wheat, and canola seed production are used in

this study. Hybrid corn seed production in China leads to emissions of 1,459.00 kg CO2e Mg-1, whereas in the U.S. it causes 2,250.50 kg CO2e Mg-1 [17,41] of emissions. Seed production of rapeseed in Poland causes slightly lower emissions of 1,014.00 kg CO2e Mg-1 [42]. In contrast, for spring wheat, it is only 580.00 kg CO2e Mg-1 in Southwest Finland and 680.00 kg CO2e Mg-1 in Northern Savonia (Finland) [43]. These emissions mostly depend on the amount of fertilizer used since fertilization is responsible for the majority of production emissions, espe- cially the production and application of nitrogen fertilizers [17]. The weighted average value used for our calculation is 1,066.26 kg CO2e Mg-1 (95% CI, 219.56, 1912.95). For the emissions of processing, packaging, and transporting cover crop seeds, only one literature source for corn seed could be found. The emissions here are 213.00 kg CO2e Mg-1 [44]. 3.11. Additional machinery operations for cover crop establishment We assume that cover crop establishment requires either tillage with a plow or cultivator fol- lowed by seeding. In addition, it is assumed that the only relevant emission from these machin- ery operations is from burning fossil diesel fuel. Based on data of an internet database from KTBL [45], plowing plus seeding is estimated at 55.00 liter diesel fuel ha-1, while cultivator till- age and seeding is estimated at 45.00 liter diesel fuel ha-1. The variability of the machinery operation intensity and corresponding fuel consumption was analyzed by Wittwer et al. [18].

In this study, diesel fuel consumption ranges from 22.00 l ha-1 for no-till systems to 61.00 l ha-1 for conventional intensive tillage systems. In organic farming, diesel fuel consumption ranges from 32.00 l ha-1 for reduced tillage to 55.00 l ha-1 for intensive tillage. The average diesel fuel consumption used in this study was calculated as 45.29 l ha-1 (95% CI, 32.43, 58.13). The emission factor per liter of diesel fuel ranges from 3.00 kg CO2e l-1 [44] to 3.31 kg CO2e l-1 [46]. The average emissions were calculated as 3.13 kg CO2e l-1 (95% CI, 2.92, 3.34) diesel fuel. The additional machinery operations thus cause emissions of 0.14 Mg CO2e ha-1 (95% CI, 0.10, 0.18). PLOS ONE Cover crops support the climate change mitigation potential of agroecosystems PLOS ONE | https://doi.org/10.1371/journal.pone.0302139 May 8, 2024 12 / 20

3.12. A factor not considered: Erosion reduction According to Lugato et al. [47], soil erosion can lead to a loss of soil carbon, lowering the car- bon sink capacity of soils. Additionally, soil erosion can disturb the soil structure and reduce soil fertility, which disconnects soil element cycles and further contributes to GHG emissions. However, the extent to which soil erosion leads to higher carbon fluxes out of the system depends on the specific context and interplay of many factors [48]. Cover crops protect soil erosion during the fall and winter and also after the establishment of the following crop, such as maize, due to remaining cover crop residues [49]. According to Laloy and Bielders [49] and Gentsch et al. [6], Cover crops increase the infiltration rate and improve soil structure, allowing the soil to absorb more water without eroding. According to Panagos et al. [15], cover crops reduce soil erosion by at least 20.0% in Europe and the United Kingdom. Machiwal et al. [50] found a reduction in soil loss of 33.0% to 77.0%, depending on the cover crop type for India. The reduction in soil erosion is highly dependent on the slope and management of the land. In addition, the soil type also plays a major role. Given the uncertainty between soil erosion and fluxes from the soil system as well as the varying factors leading to soil erosion, we have not included climate benefits of soil erosion prevention from cover crops in our calculations. 3.13. Net climate change mitigation impact (NCCMI) of cover crops

We calculate the summary of all climate benefits from cover crops as 4.23 Mg CO2e ha-1 a-1 (95%CI, 1.25, 7.49) and all climate costs from cover crops as 0.93 Mg CO2-eq ha-1 a-1 (95% CI 0.51, 1.34). The NCCMI of cover crops was calculated as 3.30 Mg CO2e ha-1 a-1 (Table 3). The NCCMI was then extrapolated to the maize harvest area for all EU-27 countries. Here, we assume a scenario in which cover crops are grown before maize on all of the EU-27’s harvested maize areas, which is 15,092 x 106 ha, on average [23]. Based on this calculation, planting cover crops before maize in the EU-27 results in a climate change mitigation potential of 49.80 mil- lion Mg CO2e a-1. This is equivalent to 13.0% of the EU-27’s agricultural GHG emissions [51]. 3.14 Sensitivity analysis based NCCMI scenarios The impact factors “carbon land benefit based on yield gain” and “carbon sequestration” have a high impact on NCCMI but also show a high uncertainty. To address this uncertainty, we use four scenarios for NCCMI to analyze the sensitivity of the results: A) Base scenario as shown in section 3.13 B) Base scenario less carbon land benefit based on yield gain C) Base scenario less carbon sequestration D) Base scenario less carbon land benefit based on yield gain and less carbon sequestration 4. Discussion 4.1. Review of results The results of this study indicate that the climate benefits of cover crops significantly exceed their climate costs. This resulted in a positive NCCMI of 3.30 Mg CO2e ha-1 a-1 that can be

achieved if cover crops were incorporated in crop rotations of maize production. The major contributor to climate benefits is through carbon land benefits based on yield gain for maize following a cover crop, which alone contributed to 33.8% of all climate benefits. Several meta studies [9,10] showed that cover crops increase soil organic carbon stocks, with an average organic carbon sequestration rate of 1.43 Mg ha-1 a-1. The most important mechanism behind this is the stimulation of microbial activity from cover crops’ organic matter input. Litter from PLOS ONE Cover crops support the climate change mitigation potential of agroecosystems PLOS ONE | https://doi.org/10.1371/journal.pone.0302139 May 8, 2024 13 / 20

shoots and roots, rhizodeposits, and the transport of photoassimilate to the rhizosphere during the growth of cover crops results in a higher and more active microbial biomass [52]. These processes resulted in microbial derived organic substances that are key to build up mineral associated organic matter fractions with prolonged turnover times [53]. The activation of microbial cycling that derived from higher organic matter input rates is therefore a strong indirect factor of the climate benefits from cover crops. Carbon sequestration was the second largest contributor (34.5%) to climate benefits. The results indicate a maize yield increase of 8.8% by following a cover crop. Since maize yield gains are higher following legume cover crops and in organic systems, it can be assumed that increased nutrient availability is one of the key factors driving crop yield gains from cover crops. Cover crops might help in closing the yield gap between organic and conventional systems, as well as in improving the resilience of arable production under climate change. However, yield benefits depend on the quality of cover crop residues and reach its maximum at litter C/N ratios <25 [54]. Closer nutrient cycling and reduction of leaching losses resulted in another 24.3% of cover crops’ climate ben- efits, including nitrogen fixation of legume cover crops. The highest maize yield gains were found for legume-based cover crops while graminoid cover crops showed only minor effects. For example, Marcillo and Miguez [24] found that a grass cover crop has no influence on

maize yield. In the same study, a legume-only mixture shows up to a 21.0% yield gain in the maize crop following the cover crop. Wittwer et al. [18] confirmed that a legume-free cover crop mixture leads to only a 3.0% increase in maize yield. Mixtures of legumes and grasses as cover crops enable yield gains from 1.3% [55] to 21.0% [56], respectively. However, cover crops can also have negative effects on the subsequent maize yield. Hunter et al. [57] showed that a high carbon to nitrogen (C/N) ratio in spring cover crops resulted in a lower yield of silage maize. Nutrient use efficiency has been demonstrated to increase through cover crops that are in crop rotations for several crops, including maize [58–60]. The inclusion of legumes increases cover crops’ litter quality and crop yield benefits; however, pure legume stands are less effective to prevent nitrogen leaching over the winter [6]. The inclusion of cover crops in crop rotations, therefore, requires a high degree of management by selection of suitable cover crop species or mixtures to maximize their green manure and environmental benefits. Cover crops are able to change the amount of incoming shortwave radiation that is reflected back to the atmosphere and thereby mitigate warming [22]. Albedo change through cover crops’ land cover has the smallest contribution (4.5%) to the climate benefits from cover crops. The systematic literature search resulted in only two publications that analyze albedo change through cover crops. Both found a slight climate change mitigating effect through radiative

forcing changes induced by an increase in surface albedo. Climate costs of cover crops were nearly six times lower than climate benefits and are mainly dominated by cover crops’ seed production emissions and foregone benefits due to cover crops’ seed land use (accounting for 41.9% and 30.1%, respectively, of the total climate costs of cover crops). Interestingly, additional machinery operations and N2O emissions from cover crops’ decomposition contributed minorly to climate costs (15.0% and 4.3%, respec- tively). Emissions of N2O have a 280 times higher global warming potential compared to CO2. Therefore, the latest studies were alarmed about the emission potential that can appear from decomposition of cover crop residues [35]. However, N2O emissions are of episodic nature and depend on extended precipitation events. Despite this, N2O emissions from cover crops’ decomposition can be further reduced if cover crop residues were not incorporated into the soil [35]. With this respect, no-till operations would further reduce the N2O emissions factor. We created four scenarios where we assumed that not all positive impacts of cover cropping might be present at the same time. We removed the major contributors: C sequestration (B) and yield gain (C) or both together (D). All scenarios kept a positive NCCMI (Fig 3). PLOS ONE Cover crops support the climate change mitigation potential of agroecosystems PLOS ONE | https://doi.org/10.1371/journal.pone.0302139 May 8, 2024 14 / 20

Therefore, we conclude that positive aspects of cover cropping always superimpose the nega- tive impacts. All scenarios show a positive NCCMI. 4.2. Comparing overall results to existing literature In contrast to Kaye and Quemada [22], this study is first to include land use effects based on the yield gain caused by cover crops. Furthermore, our study uses a systematic literature review and quantifies the climate change mitigation impact for all of EU-27, which demonstrates the magnitude of climate change mitigation opportunities of cover crops. So far no other literature has extrapolated the climate change mitigation impact of cover crops to a country or region. Overall, Kaye and Quemada [22] found a climate change mitigation effect of 1.16–1.35 Mg CO2e ha-1 a-1 from cover crops, which is lower than that calculated in this study. This discrep- ancy is a result of different analyzed literature and the inclusion of land use effects in our research. In addition, Abdalla et al. [25] showed a slightly lower climate change mitigation impact from cover crops compared to our results of 2.06 Mg CO2e ha−1 a−1(95% CI -0.04, 4.16). The study, however, does not consider land use change effects (1.51 Mg CO2e ha-1 a-1 in our study), albedo change, emissions for cover crop seed production, and additional machin- ery operations. If the study would have considered these factors, their results would be more comparable to those in our study. 4.3. Limitations A limitation of this study is that we did not include erosion reduction of cover crops as a cli-

mate benefit due to the high data variability and uncertainty of its climate impact. Further- more, the use of carbon opportunity cost as suggested by Searchinger at al. [11] is criticized by Fig 3. Sankey plots describing gains and losses of different scenarios for the NCCMI. Numbers are given as Mg CO2e ha-1. https://doi.org/10.1371/journal.pone.0302139.g003 PLOS ONE Cover crops support the climate change mitigation potential of agroecosystems PLOS ONE | https://doi.org/10.1371/journal.pone.0302139 May 8, 2024 15 / 20