Introduction

Agroforestry (AF) is gaining recognition as a tool to address the various crises faced by the agricultural sector today. It can mitigate and adapt to climate change (Nair 2011; Cary and Frey 2020), increase biodiversity (Smith et al. 2013; Graham and Nassauer 2019; Bentrup et al. 2019), and soil conservation (Nair 2011; Smith et al. 2013; Beule et al. 2019), while also improving other ecosystem services like air and water quality (Nair 2011). However, AF land use in Germany makes up less than two percent of utilized agricultural area (den Herder et al. 2017). Farmers' decisions to adopt AF and other environmentally friendly practices is often driven by their economic performance (Swinton et al. 2015; Beer and Theuvsen 2019). While AF tends to have lower economic performance than agricultural production, there are notable exceptions in areas with marginal soil and with increasing policy support (Thiesmeier and Zander 2023).

The German federal state of Brandenburg has less fertile soil and little green infrastructure. As such, AF could offer meaningful environmental benefits. While farmers may adopt AF for their environmental benefits, it remains unclear whether farmers will do so based on economic considerations, which is the focus of this study. Since there are limited experimental sites available in Brandenburg, an economic model called Agroforstrechner was used in an attempt to to assess the economic viability of switching from conventional arable to silvoarable farming in Brandenburg. To limit the scope of the paper and make use of more reliable yield data, this paper focuses on silvoarable alley cropping (AC) with poplar (see Sect. "Agroforestry systems") and aims to answer the following research questions:

  1. 1.

    How does the economic performance of poplar AC under different site conditions in Brandenburg compare to conventional arable production?

  2. 2.

    Which policy measures can compensate for potential economic losses when switching from an arable to a poplar AC system?

By answering these questions, it is possible to gain insights into the economic viability of AF in Brandenburg and inform policies to increase the adoption of AF in the future.

Materials and methods

Case study region

The federal state of Brandenburg, located in north-eastern Germany, is characterised by less fertile, sandy soils (MLUK 2020) with low water retention capacities. The region has comparatively low average annual precipitation (550mm), with most of it falling in winter. Precipitation averages around 177mm during the summer months (Deutscher Wetterdienst 2019). This, in combination with sandy soils, makes Brandenburg a relatively dry region. Additionally, the collectivisation of agricultural enterprises into large single-farm collectives during the time of the GDR resulted in the removal of many small landscape structures like hedges or tree strips. This land cover simplification has led to habitat loss and therefore decreased biodiversity (Noack et al. 2022), as well as increasing susceptibility to wind and water erosion (Weninger et al. 2021; Funk et al. 2023).

Silvoarable AC can reduce wind speed (Böhm et al. 2014), leading to reduced soil loss and evapotranspiration, making it a potential solution to increase and/or stabilise crop yields within the alleys (Kanzler et al. 2019) while providing tree biomass. Silvoarable AC can also increase biodiversity compared to cropland (Mupepele et al. 2021; Edo et al. 2024). Furthermore, there is lower opportunity cost of converting parts of already less fertile land compared to regions with higher yield potentials and an abundance of green infrastructure.

Site conditions

To differentiate between site conditions in Brandenburg, the German Soil Fertility Index (SFI) was used as a reference. It provides a measure of soil fertility based on both yield potential and land value, with scores ranging from zero (least fertile) to 100 (most fertile). For the purposes of this study, only the yield aspect was considered. Structured along the SFI, Brandenburg contains five Agricultural Zones (AZ) (Hanff and Lau 2021) which serve to evaluate the yield potential of crops and trees in different areas. The weighted average SFI for AZs 1 to 5 is: 53, 40, 31, 25, and 19 respectively.

Agroforestry systems

This paper focuses on silvoarable AF for woodchip (WC) production, i.e., the combination of woody perennials and arable crops on the same plot. While AF can also be used for nut, fruit, or timber production, such systems have higher labour requirements and cost and may not be a realistic option for many farmers. Moreover, there is no reliable yield data available for long-standing timber, fruit, or nut trees in AF systems in Brandenburg. Hence, poplar for WC production is considered the most feasible option for modelling purposes, as it has been cultivated in short rotation coppice plantations in Brandenburg for multiple years and yield data are more reliable.

In this study, AF systems are designed as alley cropping systems, which allows the operation of large-scale machinery, thereby reducing the need for additional working steps and fuel use. Wood strips are oriented in north–south direction to minimize shading. This paper uses a modelling approach with theoretical plot sizes based on Emmann et al. (2012) and model plots are 20ha in size; 330m by 606.06m. The alley width (AW) is set at 24m, 48m, and 96m. A 48m AW was chosen as it has been used on AC experimental plots in Brandenburg (Emmann et al. 2012; Böhm et al. 2014; Kanzler et al. 2019) and is based on the widest available crop protection sprayer (24m width) (Hanff and Lau 2021). An AW of 24m could be applicable for farmers who want to maximize wooded area. Wider alleys (96m) were used in an experimental design by Böhm et al. (2014) and while wind speed reduction was lower compared to more closely spaced tree strips, wind velocity was still reduced compared to open field. Alleys wider than 96m would likely not have any positive impact on microclimate and therefore growing conditions of crops. Therefore, the 96m AW would still provide microclimate benefits while minimizing the conversion of cropping to wooded area on the plot. This could be appealing to farmers interested in the environmental benefits of AF while minimising wooded area.

A 20m turning area for easy machinery use is incorporated at the end of each wood strip (Emmann et al. 2012) and the AW between the edge of the field and the first wood strip is 24m. This is in line with the minimum distance requirements stipulated in current legislation (CAP-Strategic Plan 2022). This paper focuses on AC systems with poplar in short (4-year harvesting interval) and medium rotations (8-year harvesting interval).

Short rotation alley cropping with poplar

Poplar has long been grown in short rotation (SR) plantations as well as on AF experimental plots (Emmann et al. 2012; Böhm et al. 2014; Kanzler et al. 2019). Therefore, yield data for poplar are relatively reliable and site specific (Hanff and Lau 2021).

Lamerre et al. (2015) showed that poplars SR coppices in AF systems have higher yields than in closely spaced plantations, due to the effect of edge trees. Trees benefit from increased radiation and nutrients from adjacent arable production, which leads to increased yield potential (Lamerre et al. 2015). Therefore, poplar yield in AC systems was adjusted from data for SR plantations. Reported yield increases in AC systems range from 4.4 to 43% compared to poplar trees in SR plantations (Lamerre et al. 2015; Bärwolff et al. 2016; Seserman et al. 2019; Böhm et al. 2020). To maximise this effect research recommends narrow wood strips with two tree rows (Lamerre et al. 2015; Dauber et al. 2016; Böhm et al. 2020). In practice, SR systems are usually planted in single rows with 2.4m distance to the next tree row and 50cm distance between trees within the same row. To adequately manage the soil underneath the trees while losing as little cropping area as possible to machinery undertaking maintenance in the first two years of establishment, a distance of 1.8m between tree rows and cropping area is assumed. Since both tree rows are adjacent to the arable fields a 40% yield increase for poplar in the AC systems was assumed, since yield data are based on poplar plantations. This is based on Seserman et al. (2019) and Böhm et al. (2020), which focus on AC systems in Brandenburg and therefore seem most appropriate.

In practice poplar strips may be wider and contain more than two tree rows. Our assumptions aim to maximise poplar yield in the wooded area. The wood strip is 6m wide and has a tree density of approximately 8333 trees/ha. The total share of wooded area on the plot can be found in Table 1 which varies according to AW.

Table 1 Share of wooded are on the model plot (%) and their total area (ha)
Full size table

The rotation schedule is set at 4 years with the final harvest after 24 years. Afterwards, the AF system will be removed. The tree output is WC for sale.

Medium rotation alley cropping with poplar

Alley cropping with medium rotation (MR) poplar strips can be an option for farmers who prefer greater flexibility and less frequent management and harvesting intervals. Unlike SR wood strips, farmers can delay harvest by a year or more if trees fail to reach the desired diameter, market prices are low, or when weather conditions do not permit harvesting within the limited time window stipulated by German regulations. The trees can be harvested using chainsaws, hired labour, or machinery and stored in the field until they achieve the required dryness before being marketed, transported, or shredded into WC. Typically, poplar trees from such systems are used for industrial purposes such as cellulose, pulp, WC, or even construction wood. Here, only WC for sale was considered. The design of these systems is similar to short rotation alley cropping systems. The spacing of the trees is as follows: 1.5m from the arable field, 3m between single rows, and 1.5m between trees within a single row. The wood strip has a total width of 6m and a tree density of 2222 trees/ha, allowing for additional light and space to improve the growing conditions of the poplar trees. Analogous to short rotation alley cropping (SR-AC), a 40% poplar yield increase was assumed for MR-AC. To maintain consistency with the timeframe of the SR-AC system, harvesting intervals of eight years with the final harvest in years 24 were assumed. Afterwards the wood strips will be removed. The share of wooded area was identical to the wooded area share in SR-AC system and can be found in Table 1.

Crop shares

To evaluate the economic performance of AF there needs to be comparison with conventional agricultural production without trees. To do so, the five most frequently cultivated crops in each AZ were identified (Table 2). In AZ1 to AZ4, the same five crops are cultivated on the majority of arable land. In AZ5 land is cultivated with sweet sorghum, used as temporary grassland, or lies fallow (Hanff and Lau 2021). Since these crops are generally not marketed but used on-farm, they were excluded from consideration since all crops are assumed to be sold. Using an optimisation approach to match the prevalence of each crop in each AZ as closely as possible two crop shares were generated, one with and one without silage corn. Crop shares without silage corn are termed “cash crops” (CC) and those with silage corn “fodder crops” (FC) henceforth.

Table 2 Cropping shares for each AZ (%) (E.g. winter rye is cultivated on 2.99% of arable land belonging to AZ1) (Based on Hanff and Lau (2021))
Full size table

The crop shares are meant to reflect the economic importance of these crops for overall economic performance of farms in the region, rather than real-world crop rotations and can be found in Table 3.

Table 3 Crop shares with and without silage corn for the 5 AZs in Brandenburg used for the arable and agroforestry systems. (Winter Wheat = WWH; Winter Barley = WBA; Winter Rapeseed = WRA; Winter Rye = WRY; Silage Corn = SCO; Winter Triticale = TRI)
Full size table

Crop-tree interactions

Since no biophysical modelling on tree-crop interaction is included in the Agroforstrechner, assumptions were made about the effect of trees on crop yields.

This effect depends on the arable crop, with substantial differences between winter cereals, rapeseed (C3 plants) and corn (C4 plant) (Tsonkova et al. 2012; Dufour et al. 2013; Pardon et al. 2018). For winter cereals, the competition effect is smaller than for corn, with some studies even finding yield improvement for winter cereals in AF sytems due to improvements in microclimate (Vetter et al. 2012; Rivest and Vézina 2014; Zheng et al. 2016; Kanzler and Böhm 2020), especially under dry growing conditions (Kanzler et al. 2018). Since Brandenburg is a comparatively dry region with high drought prevalence in the past years, it was assumed any negative effects would be counteracted by improvements in microclimate. Therefore, yields of CC are identical in both the AF SYSTEMS and the arable system without trees.

In Brandenburg, silage corn is cultivated on approximately 20% of arable land in any given year (Hanff and Lau 2021). Since it is a C4 plant, yields are more sensitive to reduced solar radiation (Pardon et al. 2018). Silage corn is cultivated especially on lower quality soil (Hanff and Lau 2021) where the opportunity costs of converting low-productivity fields to AF are lower. This presents a conflict of interest as many farmers may not be willing or able to fundamentally change production systems, especially if silage corn is used for animal feed or biogas. To address this, two scenarios were considered: one where improved microclimate compensates for increased shade (yield remains constant), and one where shading reduces silage corn yields. The percentage of yield reduction depends on AW (Pardon et al. 2018). Based on data for AF systems with middle-aged trees, yield depression is 30.6% in distances ranging from 2.5 to 12m from the tree rows. At a 24m AW, the yield reduction is set at 30.6%, for 48m at 15.35% (half the alley is not shaded), and for 96m at 7.65% (three-quarters of the alley not shaded) (own calculations based on Pardon et al. 2018).

Economic assessment

Agroforstrechner

The Agroforstrechner (AFR) is a tool that calculates the net present value (NPV) and annuity of arable and wooded areas on a per hectare scale. It was originally published as an MS Excel-based spreadsheet model and uses statistical data and tree yield data from expert interviews. The database was evaluated for plausibility and practical relevance by farmers (Böhm and Werwoll 2020).

The original AFR was translated into a Python-based script and a relational database to improve run time and stability. Policy payments, which were not included in the original AFR, were added, as was the ability to mix wood and crops on the same area by specifying the wooded area share per hectare. The Python based script allows for the calculation of a variety of scenarios (e.g., policy schemes, crop rotations, changing WC prices) through batch requests, making the AFR more suitable for research purposes. Prior to this evolution of the model, each individual scenario had to be calculated using a Graphical User Interface. At the time of publishing, a Graphical User Interface not yet available for the Python-based AFR, limiting accessibility for non-specialist users.

While the AFR returns both NPV and annuity for the arable, poplar, and AC systems, this paper focuses on NPV (Eq. 1) because of its wider use in AF modelling approaches (Thiesmeier and Zander 2023). Discount rate (i) (Eq. 1) was set at 4% (Giannitsopoulos et al. 2020; Kaske et al. 2021).

Lastly, crop data was updated based on Hanff and Lau (2021), as were WC prices (CARMEN e.V. 2023), and cost related to polar cultivation (Peschel 2023) (Table 4). Establishment, harvesting and management costs of the wooded area differ between AWs because of different economies of scale. The costs inquired by the arrival of machinery is constant in all AF systems but the area on which it operated differs. Therefore, AF systems with a lower share of wooded area have proportionally higher costs than AF systems with higher poplar shares.

Table 4 Data sources for the economic calculation in the AFR
Full size table

For more information on equations and input data see Appendix A and B respectively.

Equation 1 Net Present Value calculation in the AFR, with (i) as discount rate and (t) as rotation year, Costs include direct and labour costs

$$NPV = \mathop \sum \limits_{t = 1}^{t} \frac{Revenue - Costs}{{\left( {1 + i} \right)^{t} }}$$
(1)

Scenario definition

For each of the five AZs two crop shares, four silage corn yield levels, three WC prices, and five policy scenarios were developed. WC price was determined using quarterly gross prices reported by CARMEN e.V. for northern Germany over the past five years (2017–2022) (CARMEN e.V. 2023). The calculated average, maximum, and minimum prices for WC with a 35% water content are €67.51/t, €109.76/t, and €48.20/t respectively.

Policy scenarios

In the current programming period of the European Common Agricultural Policy (CAP 2023–2027), Germany introduced payments for AF under Eco-Scheme three (ES3). Originally, ES3 allocated 60 €/ha wooded area for the retention of AF, while both crop and wooded area receive direct payments (“CAP23”). As of June 2023 the German federal government has increased ES3 payments from 60€ to 200€/ha wooded area (“CAP23_max”) (BMEL 2023a). However, the German AF association has criticised ES3 payments as still being too low. Instead, they advocate for 850€/ha wooded area (Günzel 2021) (“Eco_850”). In addition to payments through ES3, investment grants for the establishment of AF are planned. The specifics are still unclear funding will not be available before 2025. Most agricultural investment aids that are currently available limit payments to 40% of the total investment sum (MLUK 2023) (“Invest_40”). However, a scenario with 100% reimbursement of establishment costs (“Invest_100”) was also included. Table 5 provides a summary of these policy scenarios and the legend for Figs. 1, 2, 3, 4. All scenarios are compared to a baseline scenario where no policy support is available.

Table 5 Description of policy scenarios
Full size table
Fig. 1
figure 1

Relative NPV of SR-AC systems with cash crops at average WC price for 24m and 48m alley widths and five AZs under the different policy scenarios. The red line symbolises the threshold at which AC becomes as profitable as arable farming. (Scenario notations see Table 5)

Full size image
Fig. 2
figure 2

Relative NPV of MR-AC systems with cash crops at average WC price for the 24m alley width and five AZs under the different policy scenarios. The red line symbolises the threshold at which AC becomes as profitable as arable farming. (Scenario notations see Table 5)

Full size image
Fig. 3
figure 3

Relative NPV of SR-AC with cash crops at 24m alley width under three different price scenarios: average (avrP), maximum (maxP), and minimum (minP) prices. The red line symbolises the threshold at which AC becomes as profitable as arable farming

Full size image
Fig. 4
figure 4

Economic performance of SR-AC systems with fodder crops and yield reduction in the corn component compared to arable farming for all alley widths and AZs. The red line symbolises the threshold at which MR-AC is as profitable as arable farming (Scenario notations see Table 5)

Full size image

Results

Alley cropping with cash crops at average WC price

All SR-AC systems with cash crops at average WC price (Fig. 1) are not competitive with arable farming without policy support. Figure 1 shows that the lower the wooded area share of a system is, the closer are the NPVs of SR-AC systems to arable farming. This holds true for the 96m alley width, too, where NVP is even closer to that of arable farming. Under scenarios CAP23, CAP23_max, and Invest_40 this situation doesn’t change. Invest_100 shows more favourable results and SR-AC becomes competitive in all AZs at all AWs. However, the economic advantage becomes especially small for AC with lower wood shares in AZs with higher yield potential. Under those conditions, the increase of NPV from transitioning to AF was very low (less than 1%). Due to differences in cost structure between the three AWs, it is seems preferable to choose a higher wooded area share due to economies of scale.

When comparing SR and MR-AC, the pattern remains similar, although NPVs of MR-AC are lower than for SR-AC. Therefore, the policy payments necessary to make MR-AC competitive with arable farming have to be higher. Only scenario Eco_850 is able to make MR-AC competitive in all AZs and for all AWs.

Alley cropping with fodder crops at average WC price

At average WC price and without corn yield reduction, SR-AC systems with corn-dominated crop shares are uncompetitive with arable farming without policy support (Fig. 2) in the Baseline, CAP23, and CAP23_max scenarios. This is in line with the results for SR-AC with cash crops. Scenario Invest_40 allows SR-AC to become competitive in AZs 3 through 5 for 24m AW, AZs 3 and 5 for 48m AW, and AZ 5 at 96m AW. Under scenario Invest_100, SR-AC with fodder crops becomes competitive in all AZs and AWs, as does Eco_850, which increases the advantage of AF over arable farming considerably.

MR-AC has less favourable economic outcomes than SR-AC as the reduction in NPV is higher. Scenario Invest_40 is insufficient to make MR-AC competitive for any AZ or AW. Scenario Invest_100 is necessary for MR-AC to be competitive in AZs 3 and 5 for all AWs. Under Eco_850, MR-AC becomes competitive for all AZs at 24m AW but remains uncompetitive in AZ 1 for the 48 and 96m systems.

Sensitivity analysis

Woodchip prices

To understand the effect of WC prices on the competitiveness of AC systems two more price levels are considered: maximum (€109.76/t) and minimum WC prices (€48.20/t) (Fig. 3).

At maximum WC price, the profitability of AC improves drastically. For SR-AC with cash crops, subsidies are no longer needed in AZ1 and AZ3 at 24 and 48m AW. For the other AZs, as well as the 96m system, policy payments can be considerably lower than at average WC price to ensure competitiveness. Scenario CAP23_max is sufficient to bridge the gap between arable farming and SR-AC and allows all AWs to become competitive (except AZ 4 at 96m AW) (Fig. 3, “maxP”). For MR-AC with cash-crops high WC price are insufficient for MR-AC systems to become competitive with arable farming and policy payments are still necessary. This is due to the generally lower NPVs for MR-AC compared to SR-AC. Scenario CAP23_max is able to make some AZs and AWs competitive. However, to make MR-AC systems competitive for all AZs and AWs, Invest_100 is needed.

At low WC prices the NPV of AC compared to arable farming drastically decreases. For SR-AC with cash crops, scenario Invest_100 is able to facilitate competitive in AZ 5. In all other AZs, Eco_850 is needed to make SR-AC competitive. While results differ between AWs, Fig. 3 (“minP”) depicts the overall pattern of distribution for the sensitivity analysis, in which economic performance increases and falls alongside the respective WC price. For MR-AC, Eco_850 is the only scenario under which MR-AC with cash crops became competitive, but only in some AZs (MR-AC remains uncompetitive in AZ1 at 48 and 96m AW and in AZ 2 at 96m AW).

AC with fodder crops (no corn yield reduction) reacts similarly to AC with cash crops but NPVs are lower. SR-AC with fodder crops at high WC price is able to be competitive without policy support in AZs 1, 3 and 5 at 24 and 48m AW. At 96m AW, SR-AC is only competitive in AZ 3 without policy support. Predictably, policy payments to make SR-AC competitive can be lower when the WC price is high. Scenario CAP23_max is sufficient to ensure competitiveness for all AZs and AWs. At low WC prices, Invest_100 is needed to make SR-AC competitive in AZs 3 through 5, and Eco_850 ensures competitiveness in all AZs.

For MR-AC with fodder crops, a high WC price is unable to make these systems competitive without additional policy support. Here, Invest_40 allows all AZs and AWs to become competitive (Except AZ 4 at 96m AW). A subset of AZs becomes profitable already at lower policy payments. At 24m AW CAP23_max is sufficient to make all except AZ4 profitable. At low WC price, MR-AC is uncompetitive except under scenario Eco_850, where all AZs and AWs become competitive.

Corn yield reduction

Yield reduction in the corn component drastically reduces competitiveness of AC with arable farming. At average WC price for SR and MR-AC, no AZs and AWs are competitive, with and without policy support. At high WC price, SR-AC remains uncompetitive except under scenario Eco_850, which made SR-AC competitive for all AZs and AWs. MR-AC was only competitive in some AZs under scenario Eco_850. For the 24 and 48m systems competiveness was limited to AZ 3. At 96m AW none of the policy scenarios make MR-AC competitive.

At low WC price no policy scenarios were able to make AC profitable for any AZ or AW, both for the SR and MR-AC systems. In this study, policy payment are unable to incentivise AF for corn dominated crop shares if yield losses are accounted for, unless WC prices are high.

Discussion

Results of this study align with findings of comparative European studies on agroforestry economics in that prices of tree products, e.g. fruit, nuts, or timber, is one of the deciding factors in profitability (Thiesmeier and Zander 2023). When prices are low, most systems are only competitive with policy support and even high tree product prices do not guarantee competitiveness. Furthermore, the gap in profitability between silvoarable AC and conventional arable farming seems to be lower on more marginal soils since opportunity costs of forgoing some cropping area in favour of trees do not impact profitability as much as on land with higher yield potential (Thiesmeier and Zander 2023).

Regarding rotation schedules, SR-AC was more economically competitive with arable farming than MR-AC. However, MR-AC could still be attractive if farmers prefer more flexibility in the harvesting schedule. For SR-AC the rotations are relatively fixed since harvesting usually involves a converted forage harvester that can only operate up to a certain tree diameter. Due to planting density the removal of single trees is impractical. In MR-AC systems plating density is lower and harvesting machinery usually includes chain saws or forestry machinery, which are more flexible when it comes to tree diameter. It has been shown that flexibility can be an important parameter for the willingness to adopt AF (Frey et al. 2013). While MR-AC is less flexible than annual cropping it is more flexible than SR-AC. However, cash flow-characteristics favour SR-AC since they generate returns sooner than MR-AC. It is therefore difficult to identify a best-practice system since they depend on farmer preference, site conditions, and machinery endowment.

Additionally, economies of scale have a sizable impact on the cost structure of AF systems. For smaller quantities of trees, relative transport and travel cost of machinery may be very high, which can make AF unprofitable. However, smaller plots or lower tree density can still be viable if AF systems from other farms are nearby and costs are shared. Also dependent on scale are management and establishment costs. On smaller plots, longer, and therefore more expensive, poplar rods would be used as planting material to limit wildlife damage. While wildlife damage still occurs on larger plots, losses would likely be limited to the field edges and proportionally smaller. Thus, shorter rods can be used on larger plots as the amount of seedlings could make up for wildlife or other losses.

Policy Implications

The German CAP-strategic plan states agroforestry will contribute to three goals, which are mainly concerned with: climate change adaptation, mitigation and protection; sustainable development and efficient resource use (air, water, soil); and containment and reversal of biodiversity loss. More specifically, AF contributes to meeting specific needs identified in the strategic plan, which are ascribed a high or very high priority (BMEL 2024). They are:

  • Securing and improving carbon storage and sequestration

  • Adapting agriculture and forestry to climate change

  • Reduction of water consumption with regard to the soil and landscape water balance

  • Protection and sustainable use of biodiversity

Currently, only Eco-Scheme payments for maintaining agroforestry systems are available (ES3; €200/ha wooded area). While investment funding will be introduced through the Common Task “Improvement of agricultural structures and coastal protection” (German: GAK), there is not information on when it will start. In the GAK plan, establishing SR systems will be subsidised with €1566/ha wooded area (BMEL 2024). Another funding opportunity will become available through the “Action Program Natural Climate Protection” from the Federal Ministry for the Environment (BMUV), which will fund the establishment of additional landscape elements, including agroforestry systems. What these measures will look like is yet to be determined (BMUV 2023).

Although there are multiple future funding opportunities, this study shows that current policy payments (ES3) are insufficient to financially incentivise AF over a wide range of site conditions and WC prices. Therefore, re-negotiating ES3 payments would be justified, especially since only 51 ha of wooded area within AF systems were registered under ES3 in Germany in 2023 (2023b; Lehmann 2023). While payment levels stated in the GAK plan are an improvement to the previous situation (i.e. no funding), they are insufficient to incentivise the establishment of AC systems. In this study, investment costs for short and medium rotation systems ranged from €4884/ha to €6322/ha wooded area (Appendix B). Therefore, the GAK payments are below the 40% reimbursement ratio in the Invest_40 scenario. Since this scenario was unable to make most AC systems competitive, lower payment levels will inevitably fail to incentivise short and medium rotation systems. However, GAK payments are higher for more complex and longer-standing systems (e.g. high-value timber or fruit trees) and could incentivise these systems. Furthermore, the classification of MR-AC is unclear.

Unfortunately, the budget for ES3 has been cut drastically (by 75%) and targets for establishing wooded areas within agroforestry reduced from 175000 to 11500ha wooded area within AF systems (Günzel 2024). Therefore, AF seems to remain a comparatively low priority within the agricultural policy landscape. This is at odds with agroforestry being listed in combination with high and very high priority needs in the strategic plan (BMEL 2024). It should also be considered that sustainability measure and Eco-Schemes can compete amongst themselves, especially in more marginal areas. From 2024 onwards, every German farm will be required to establish fallows on 4% of its surface area (GAPKondG 2022). Since these 4% will not be financially compensated they will likely be established on plots with lower yield potential to limit economic losses. To offset some of those losses, farmers will likely cultivate additional fallow under ES1. Since those payments are much higher than for ES3 and require less investment and overhead costs, AF seems to be at a disadvantage.

Methodological reflections

The sensitivity analysis for corn yield revealed that relative yield is a key factor for economic competitiveness. If AF leads to corn yield reduction the economic performance decreases drastically. However, assumptions are based on a Belgian study because no data is available for Brandenburg. Since SR trees are harvested regularly, the effect on corn yields could be lower, especially if its cultivation coincides with the tree harvest when shade would be limited. Microclimatic effects and reduced wind velocity (Böhm et al. 2014; Kanzler et al. 2019) could also benefit corn intensive AC systems. In water limited regions like Brandenburg, yield stability could be improved, especially when trees fix nitrate and therefore provide additional fertilising benefits (Sileshi et al. 2012; Nasielski et al. 2015). If these positive impacts can make up for the shading effect, AF with corn could still be an economically viable land use option.

Additionally, it remains unclear how crop-tree interactions will behave under changing climatic conditions. There are challenges in projecting the future yield performance of crops in AF systems (Luedeling et al. 2014) and no such projections are available for Germany. However, literature suggests that AF can buffer crops from climate change and volatile weather patterns and stabilise yields (Quandt et al. 2023). It is also possible that assumptions on cash crop yields could be too optimistic in this study. If they decrease through the establishment of agroforestry (Ivezić et al. 2021), this would decrease the competitiveness with arable farming and increase the amount of policy payments needed.

WC prices also have a strong influence on the competitiveness of AC systems. While prices have increased in recent years, arguably due to the energy crisis brought on by Russia’s invasion of Ukraine, prices have fluctuated (CARMEN e.V. 2023). It is unclear how they will behave in the future and how oil prices will affect prices for alternative energy sources. Due to the long life-span of AF systems, competitiveness will likely be affected by a range of WC prices at different times within the rotation. This is not reflected in this paper.

Input data is taken from statistical handbooks and websites and it is unclear how prices, costs, and growing conditions might change in the future. This makes predictions for long standing systems difficult. Therefore, our results show the economic performance under “business-as-usual” conditions where prices, costs and yields remain constant over the lifespan of the AF system.

The only marketing route considered in this study is the sale of WCs. However, other value chains might be relevant, e.g., industrial use or timber production. This may result in diverging harvest costs and/or product prices. WC can also be used on-farm which would necessitate a different approach for assessing economic performance. However, this is beyond the scope of this study.

The Agroforstrechner only includes market costs for AC systems. Additional costs to farmers such as e.g., (self) education, system design and management are not included but can be substantial and therefore a barrier to AF adoption. Additionally, the environmental benefits of AC systems are not included in market prices, only their indirect effect on crop yields is reflected.

Outlook for future research

A major limitation for the scope of this paper is data availability, especially for non-poplar trees in agroforestry systems (e.g. high value timber, nut, or fruit trees) in Germany. Therefore, this author recommends data collection for existing systems as well as establishing new experimental sites. These could also provide more insight into tree-crop interactions, especially with C4-plants. While primary data is important it is also key to consider more modelling approaches for understanding tree-crop dynamics and yield performances under climate change. These kinds of projections are currently unavailable for Germany. Research could inform farmers and policy makers on long-term benefits (or disadvantage) in terms of yield stability and protection from climate change effects.

Limited research is available also on value chains for AF products. The economic potential of non-poplar, non-fuel wood systems is still scarcely documented.

In terms of policy, there is a need to better understand the trade-offs between different policy measures and how they compete amongst themselves. While there are some studies evaluating effectiveness of policy measures (Wiegmann et al. 2022) their economic impact on farm level is lacking. While this paper found a certain volatility in terms of economic performance dependent on price and subsidy levels, both yield and economic risk of AF systems is not widely researched.

Conclusion

Recent policy efforts have been made to support AF in Germany. However, the economic viability of such systems in comparison to arable farming remains unclear, with and without policy support. As profitability is an important driver for farmer adoption, this study looked at the economic potential of AC systems in Brandenburg.

This paper has found that without policy support, high WC prices are necessary for AC to become competitiveness and even then, only some SR-AC are able to do so. At average WC prices however, no systems are competitive without policy support. Generally, SR-AC had higher NPVs than MR-AC due to the cash-flow structure of these systems but farmer may have other preferences, e.g., flexibility.

To ensure competitiveness with arable farming policy support is needed. However, current support offered under ES3 is insufficient to make AC competitive over a wide range of soil conditions and WC prices. The introduction of investment aid for the establishment of AF systems improves this but payment levels are crucial. A full coverage of investment costs would make all AC systems competitive at high WC prices, regardless of AW and AZ. At average WC prices this holds true only or for SR-AC without yield in the crop component. MR-AC requires higher ES3 payments of 850€/ha wooded area in order for all AZs and AWs to become competitive. When yield loss in corn is included, nearly all SR and MR-AC were unprofitable, even at high WC prices.

In summary, without policy support AC was scarcely competitive with arable farming and both current (ES3) and proposed (GAK plan) policy payments are insufficient to incentivise AC. Financial incentives for AF in Germany seem disjointed (ranging from EU to federal level), at odds with the priorities in the strategic plan, and do not match the real costs of these systems. Increasing investment grants and/or ES3 payments would be necessary but seem implausible given the current agricultural policy landscape. As such, the full potential of AF in Germany is unlikely to be realised in the coming years.