Reduction of the economic risk by adaptation measures to alleviate heat stress in confined buildings for growing-fattening pigs modelled by a projection for Central Europe in 2030

: Economic risks for livestock production are caused by volatile commodities and market conditions, but also by environmental drivers like increasing uncertainties due to weather anoma-lies and global warming. These risks impact the gross margin of farmers and can stimulated investment decisions. For confined pig and poultry production, farmers can reduce the environmental impact by implementing specific adaptation measures to reduce heat stress. A simulation model driven by meteorological data was used to calculate heat stress impact as a projection for 2030. For a business-as-usual livestock building, the indoor climate for several adaptation measures was calculated. The weather-related value-at risk quantified the economic risks caused by global warming and the stochastic component of the weather. The results show that only energy-saving adaptation measures to reduce the inlet air temperature are appropriate to reduce the economic risk to the level of the year 1980. The efficiency of other adaptation measures to reduce heat stress is distinctly lower. The results in this study can support the decision making of farmers concerning adaptation management and investments. It can inform agricultural policy design as well as technological development.

In mid-latitudes the majority of pigs and poultry are kept inside confined livestock buildings, which are often perceived as elements of industrial livestock systems [11,12]. The predominant type of confined livestock buildings is characterised by a mechanical ventilation system, high stocking density in insulated buildings. At global level, more than half of the pig and poultry populations are kept under these husbandry conditions [13]. Consequently, HS inside livestock buildings increased even in a temperate climate of Central Europe in the range between 0.9 and 6.4% per year between 1981 and 2017 [14]. This has an impact on the economic risk (ER) for livestock production, which may further increase in the future. The communication of the economic impact of climate change in the context of livestock farming is often addressing the national level. In France and Spain, the extremely hot year 2003 caused additional costs of 100 million € each for the intensive livestock sector and a feed deficit in the range of 30% for Germany, Austria and Spain, about 40% in Italy and about 60% in France. In Spain, the mortality of broilers was in the range of 15 to 20% with a decrease of productivity by 25-30%. In France, about 4 million broilers died [2]. For fattening pigs, Schauberger et al. [15] determined the loss in gross margin for a 10-year return period in an Austrian case study with 0.27 € per year and animal in 1980 and a 20-fold increase to 5.13 € per year and animal in 2030. These economic losses are caused by an increase of the feed conversion ratio, a lower growth rate, and other productivity traits like increasing mortality [3,[16][17][18][19].
Adaptation measures (AMs) have been proven effective to reduce HS in confined livestock buildings [15,20,21]. They can be grouped into measures adapting (1) the ventilation system to modify the thermodynamic properties of the inlet air (air temperature and humidity), (2) the building (e.g., insulation, orientation) [22], (3) the indoor equipment at the animal level to modify the indoor climate on a small scale (e.g., sprinkling and fogging systems, forced ventilation), and (4) the livestock management (e.g., thermotolerant breeds, feeding, design values of the ventilation system) [20].
The scale of the operating figures is beyond everyday experiences of farmers. Neither the reported HS parameters for animals [14] nor the projections for the near future [15] can be directly used by farmers to adopt AMs. Livestock specific information on the likelihood and severity of future HS and the costs and benefits of AMs support investment decisions in new livestock housing systems and reduce the economic risks of stranded assets in the field of livestock farming. In addition, all future approaches to increase sustainability in animal production systems have to target the alleviation of HS. Besides animal welfare improvements, possible economic benefits are important arguments for farmers to adopt AMs and thereby manage the economic risks caused by HS.
Due to the stochastic element of the weather, the concept of the weather-related Value-at-Risk (weather-VaR) was applied as a decision tool [23,24]. We use the weather-VaR to describe the ER caused by HS on the gross margin of growing-fattening pigs. With this tool, the expected mean value of the reduction of the gross margin caused by global warming can be calculated but also the return period (10 or 20 years), taking into account the stochastic weather behaviour [15]. This paper describes a three-step process. In the first step, the likelihood of HS is calculated using a simulation model for indoor air of a confined livestock building, which is driven by meteorological data. The stochastic behaviour of weather elements is the primary input for the uncertainty from year to year. The reduction of HS by several AMs is calculated in comparison to a business-as-usual pig building as a reference. In a second step, the impact of HS on the gross margin for the fattening pig sector is calculated by the modified impact function of St. Pierre et al. [17]. In the last step, the likelihood of HS and the impact function are combined to determine the likelihood of the ER for the near future (i.e. 2030) on the basis of the weather-VaR concept for the reduction of the gross margin per animal place [15,24]. The reduction of the ER due to the application of AMs is the relevant output to support farmers' management decisions for the planning of new or the adaptation of existing livestock buildings.
The following questions will be answered by this investigation: (1) which AMs are appropriate to reduce HS and to increase the resilience of livestock buildings against global warming, (2) what is the expectation for the occurrence of HS in 2030 by the use of AMs, and (3) what is the expectation for the mean ER and the VaR in 2030.

Simulation of the thermal climate inside livestock buildings
The likelihood of HS is determined by a simulation model of the indoor climate of a confined livestock building, which is driven by meteorological data, directly measured at the site of investigation, the city of Wels in Upper Austria (48.16°N, 14.07°E) for the period 1981 to 2017 with a temporal resolution of one hour.
The indoor climate of the livestock building was simulated by a steady-state model, which calculates the indoor parameters, air temperature and humidity, and the ventilation flow rate. The thermal environment inside the building depends on the livestock, the thermal properties of the building, the ventilation system, and its control unit. The core of the model is the calculation of the sensible heat balance of a livestock building [14,25,26]. The model calculation was performed for a typical building for fattening pigs in Central Europe for 1800 heads, divided into nine sections, with 200 animals each. The model calculations were performed for the entire growing-fattening period, i.e. for a body mass between 30 and 120 kg. The system parameters (livestock, thermal properties of the building and ventilation system) of the livestock building are discussed in detail in [14]. The HS of the business-as-usual system REF was compared afterwards based on the HS index, which is reduced by the application of different AMs.

Quantification of heat stress
The temperature-humidity index (THI) was used to quantify the indoor HS for pigs according to THI = 0.72 T + 0.72 TWB + 40.6, where TWB represents the wet-bulb temperature. The THI threshold X = 75 represents an alert situation of the thermal environment [22]. The intensity of HS, ATHI75, was assessed by the aggregated values larger than X = 75 and the course of time of the THI (i.e. the area under the curve over time). The THI was calculated for the reference system (REF) and the seven modelled AMs to reduce HS. The THI index was further processed as annual sums.
The expected value (mean value) of the THI for the modelled AMs for a certain year t, was calculated by a simple linear model, fitted to the logarithmically transformed yearly values ATHI75 by mt = k t + d, with the slope k and the intercept d. The linear regression was evaluated by the correlation coefficient r and the corresponding p-value.
The logarithmically transformed values were detrended by a linear regression according to t = mt -log ATHI75,t. The expected occurrence of the HS index is calculated by a normal distribution N(mt., ²), using the mean value from the linear regression mt of a certain year t and the variance ² of the detrended values t [15]. The projection for the near future was performed for the year t = 2030.
The trends of the log ATHI75 for the modelled AMs were analysed by the signal-tonoise ratio (SNR) and the Mann-Kendall Trend Test. The SNR was calculated using the linear trend k over 37 years and the standard deviation of log ATHI75. Under the assumption that the SNR is distributed with the standard normal distribution N (0; 1), limits for the SNR and the p-values are as follows [27]: low significance 1.645 < SNR ≤ 1.960 (0.05 < p ≤ 0.10), medium significance 1.960 < SNR ≤ 2.576 (0.01 < p ≤ 0.05), and high significance SNR > 2.576 (p ≤ 0.01). The second test for the trend of the HS indices was the rankbased non-parametric Mann-Kendall trend test with the test statistics τ and a one-sided test for an increasing temporal trend using the R package trend.

Modification of HS inside pig buildings by the use of adaptation measures
For this investigation, two groups of AMs were analysed. The efficacy of the first seven AMs was assessed by the simulation model [21]. The efficacy of the second group of 14 AMs was estimated by a panel of experts using the modelled AMs as a reference [20].
In the first group the following seven AMs were modelled: The first three are energysaving systems which are part of the ventilation system and modify the thermodynamic properties of the inlet air (air temperature and humidity) [22]. Three different energy-saving systems were investigated: (1) direct evaporative cooling by cooling pads (CP), (2) an indirect evaporative cooling by the combination of CP with a regenerative heat exchanger (CPHE) to reduce the humidity load caused by the evaporative cooling, and (3) an earthair heat exchanger (EAHE), using the ground as heat storage. The fourth AM assumes an inversion of the feeding and resting times by half a day to move the maximum heat release of the animals to night-time with a lower inlet air temperature INV. Two further AMs modify the internal heat load of the livestock building by a reduction of the stocking density SD to (5) 80% of the design value (SD80%) and (6) to 60% (SD60%). The last AM (7) affects the design value of the ventilation system by doubling the maximum volume flow rate (VENT) to increase the removal of the sensible heat released by the animals.
An assessment of the estimated efficacy of the following AMs by an expert panel was performed [20]:

Economic impact of HS
The economic impact assessment is based on the reduction of the gross margin as a function of HS, estimated by the growing-fattening pig model of St-Pierre et al. [17]. The reduction of the gross margin is calculated by three parameters on an annual basis for one animal place: the reduction of body mass at the end of the fattening period (kg a --1 ) (revenue), the reduction of dry matter intake (kg a --1 ) (variable costs) and the increase of mortality (%) (revenue). These three parameters were updated by data for 2020 from the Federal Institute of Agricultural Economics, Rural and Mountain Research (BAB http://www.BAB.gv.at/) for feed with 0.25 € kg --1 , the revenue for a fattening pig with 1.7 € kg --1 , and the cost of a dead pig (75 kg) with 100 € [15]. The original predictor of the impact function for growing-fattening pigs of St-Pierre et al. [17] is the area under the curve ATHI75 for a THI threshold of X = 72. By the dataset of the HS index using the two THI thresholds, X = 72 (growing-fattening pigs) and X = 74 (sows), the impact function was modified for the THI threshold X =75 used in this paper.
The economic impact IMP (€ a --1 ) related to one animal place as a linear function of the HS index ATHI75 (h a --1 ) reads as follows: IMP = 0.0016 ATHI75 with r² = 0.9897 ( Figure 1).

HS of the reference livestock building (REF) and the reduction of HS by the application of AMs
The relationship of HS between the reference livestock building (REF), which is calculated for a business-as-usual management, and the modelled AMs is shown in Figure  2. The lowest performance is reached by a reduction of the livestock density (SD80% and SD60%), followed by inverting the activity of the animals to the night-time (INV) and by doubling the maximum volume flow rate (VENT). The distinctly best performance can be expected for energy-saving systems, which cool the inlet air (CPHE, CP, and EAHE). All investigated AMs show a lower HS inside the livestock building compared to the REF building.  [20].
The resilience of the confined livestock system against HS was assessed by the relationship between the HS indices ATHI75 calculated from the outdoor parameters on the one hand and the indoor climate for the reference system REF and buildings equipped with one of the seven modelled AMs. The linear regression of the outdoor situation on the indoor situation is shown in Figure 3. If the HS indices are above the line of identity, then the resilience of the indoor climate is lower compared to the outdoor situation (REF, SD80%, SD60%, INV, and VENT). Only the AMs, which reduce the inlet air temperature, show a higher resilience compared to the outside situation. For these AMs, the indoor and the outdoor situation is decoupled to a large extent. The lower resilience results in a higher HS indoors compared to the outdoor situation (above the line of identity between HS outdoor and indoor).  The temporal trend between 1981 and 2017 of the HS index ATHI75 was analysed by a linear model ( Table 1). The logarithmically transformed HS parameter shows a highly significant correlation (p < 0.01) for the reference building REF and for the seven modelled AMs. The temporal trend was tested by the signal-to-noise ratio SNR and the Mann-Kendall Trend Test (Table 2). Both tests show a significant trend, at least with p < 0.05. The higher the efficacy of the AMs, the lower the statistical significance.

Consequence of global warming on the likelihood of heat stress
The parsimonious model, using the linear regression of the HS index ATHI75 was used to determine the likelihood of its occurrence in a certain year t. In Figure 4

Economic Risk for Pig Farms due to Global Warming
The economic risk ER due to a global warming trend for the pig fattening sector was assessed by the product of the likelihood of HS (Figure 4) and the economic impact function IMP, which describes the expected reduction of the gross margin as a function of the HS index ATHI75, shown in Figure 1. The probability density function PDF of the ER is presented in Figure 5. The expected value of the ER for the near future, i.e. 2030, is shown by the maximum of the PDF. The shift of the mean value to higher values is caused by the global warming trend, whereas the uncertainty (PDF) is caused by the stochastic character of the meteorological data. This is represented by the width of the PDF in Figure 5. All AMs reduce the ER but the three AMs which reduce the inlet air temperature are showing by far the best performance. The weather-VaR is represented by the 10 a return period (90 th percentile) or the 20 a return period (95 th percentile). These two parameters can be read directly from the cumulative distribution function CDF of the ER, shown for the seven AMs and the REF in Figure 6. The CDFs for the air treatment AMs are located below a gross margin reduction of 1 € a --1 per animal place, in the same range or even better as the reference building REF in 1980. The detailed data for the expected value and the two return periods are summarised in Table 3. Table 1. Statistics of the economic risk by the reduction of the gross margin (€ a -1 ) per animal place for the year 2030 by the expected values of the economic risk (mean) and the weather-VaR for a return period of 10 a and 20 a, respectively. Besides the seven modelled AMs (e. g., Table 2), an assessment was included of the estimated efficacy of the fourteen additional AMs by an expert panel was included [20] as mentioned above.

Reduction of the Gross Margin
For these AMs, the reduction factor was estimated based on the experience from experts in the field of agricultural engineering and veterinary medicine. Experts were asked for the expected minimum and maximum values for heat stress reduction by a single AM compared to a reference situation. These values led to an min. and max. reduction factors RFAM. The HS index of an estimated AM ATHI75,AM is then calculated by multiplying the estimated reduction factor RFAM and the HS for the reference building REF  Table 4, showing the mean value of the ER and the confidence interval for the modelled AMs and the range (maximum-minimum) of ER for those AMs, which were evaluated by the expert panel.
The ER are lowest for the energy-saving air treatment systems and highest for building features. For these features, the reduction factors were less than 8%. The AMs, which are working on an animal level, modifying the microenvironment of the animals, show a wide variety of efficacy. For all AMs, the presented reduction factors and the corresponding ER assessment are only valid for the application as single AM. Table 4. Overview of the investigated adaptation measures AMs, the method of the efficacy assessment (model result M / expert estimate E), the efficacy (%) in reduction of the heat stress parameter ATHI75 compared to reference system REF (the estimated efficacy is given as range (in brackets) [15]. The projection of the economic risk is given by the expected value (mean value) and the confidence interval [upper limit, lower limit] or by the estimated range of the efficacy (maximum, minimum).

Adaptation measure Method
Efficacy (%) in reduction of the heat stress index ATHI75 compared to reference system REF

Discussion
Global warming is also strongly influencing Central Europe. This is not only recognizable in the annual values but also in seasonal patterns of core climate indicators. The increasing frequency of hot days and heatwaves in recent decades due to global warming has increased the likelihood for situations that cause HS [28][29][30][31][32][33]. The prediction of HS in the near future due to global warming and the uncertainty caused by the stochastic behaviour of the meteorological data index can be quantified by a probability density function [33], showing an increasing threat for confined livestock farms [15]. For a Central European case study [15], the simulation of the indoor climate for fattening pigs shows a distinct increase for several HS parameters. For example, the frequency of the exceedance of an indoor temperature threshold of 25°C increased between 1989 and 2020 from 7.1% to 10.5%. For the THI threshold of 75, the frequency more than doubled from 2.8% in 1980 to 6.3% in 2020.Schauberger et al. [21] used the HS index ATHI75 (area under the curve A and a THI threshold of 75), which increased from 449 in 1980 to 1666 in 2020, by 271%. Schauberger et al., [15] extended this analysis by a short-range projection for 2030 extrapolating the linear trend over 37 years. Here, the linear trend for a reference livestock building and a number of distinct AMs was calculated using the logarithmically transformed HS index ATHI75. The extrapolation of the trend can be expected as a robust prediction [34]. The trend projections for 2030 have the advantage that not only the expected value (mean value) of the HS index ATHI75 was determined, but also the uncertainty under the assumption of a log-normal distribution. The assumption of linear trends holds for the short-range period in this study but may lose ground for long term assessments.
Sutton [35] emphasises that assessment and communication of risk merits greater attention. Furthermore, climate change adaptation has become mainstream in agricultural policy agendas. This means that climate change should be framed as a problem of risk assessment and risk management. Therefore, the likelihood of HS was multiplied by an impact function to quantify the risk by HS. This approach was applied for fattening pigs, using the reduction of the gross margin as impact function to quantify the ER. The impact function IMP (Figure 1) describes the relationship between the gross margin reduction for one animal place and the HS index ATHI75 [17], which has been adapted to the case study situation and the reference year 2020 [15]. For a business-as-usual livestock building REF, the mean reduction of the gross margin grew from 0.08 € a −1 to 1.57 € a −1 per animal place for the period from 1980 to 2020. Taking into account the likelihood of the occurrence of HS, ER was determined for a return period of 20 years with 0.38 and 7.18 € a −1 per animal place for 1980 and 2030, respectively [15].
This temporal trend shows that, especially for new buildings with an operating period of several decades, there is an urgent need to consider the implementation of AMs to tackle the HS challenge. Therefore, we investigated AMs with a wide variety of efficacy to reduce HS. Data is based on an expert survey of AMs that resulted in HS reduction factors [20], as well as on a simulation of seven AMs [21]. The best performance can be expected for systems that cool the inlet air (CP, CPHE, GEO, and EAHE) with the highest reduction factor for EAHE and GEO [22]. The lowest performance can be expected for measures concerning the features of the building (e.g. ORIENTATION). The efficacy of AMs at animal level (e.g., SPRINKLING, FOGGING) and concerning livestock management (e.g., reduction of stocking density SD80%, SD60%) show a medium performance with remarkable variability. Besides the reduction factor as a quantitative measure of the AM efficacy compared to a reference building also the improvement of the resilience by the application of the AMs was investigated. The reference building REF shows a lower resilience for global warming compared to the outside situation [14]. In comparison to the situation indoor of the reference building REF, all modelled AMs show a greater resilience ( Figure 2). The model calculations also showed that the resilience for those AMs, which cool the inlet air, is even higher in relation to the outside situation ( Figure 3). This decoupling of some modelled AMs from outside weather conditions was proven by a higher statistical uncertainty (p-value) of the temporal trend ( Table 2).
The economic risk (ER) for a reduction of the gross margin shows a similar behaviour ( Figure 5). For the CP and EAHE, the mean value of the ER is lower compared to the risk of REF for 1980. This means that, by using these AMs, the impact of global warming can be largely eliminated for the short range period until 2030. Whether this holds for long range periods with stronger climate change signals needs further investigations. The CPHE shows a slightly higher ER. For the two management related AMs, VENT (doubling of the ventilation rate) and INV (inverting of the resting and activity period), a better performance with a lower ER can be expected compared to a reduction of the stocking density (SD80% and SD60%). Using the concept of weather-VaR, not only the expected value (mean value) is taken into account but also the uncertainty by the 90 th percentile (10 a return period) and 95 th percentile (20 a return period). For these values, the improvement by the use of AMs is not very high with the anticipated result, that the AMs as part of the ventilation system (cooling the inlet air) show the best performance.
The reduction of the ER with AMs is likely accompanied with additional operating costs including farm labour demand. Furthermore, investment costs have to be taken into account as well, which are typically evaluated by farmers in relation to the profitability of the farm operating with a status-quo reference building. Operating and investment costs lie in a wide interval, depending on the farm size (economies of scale), labour cost or input and output markets (e.g., prices for feed, piglets, and meat) (Germany [36], EU: production costs [37,38], Austria: Federal Institute of Agricultural Economics AWI www.awi.bmnt.gv.at; US: investment costs [39] and variable costs [17], Belgium [40]). The operating costs of AMs are determined by electricity demand (e.g., water pumps for FOG-GING, SPRINKLING, CONDUCT, RADIANT; pressure drop for EAHR, CP, CPHE, GEO, FORCED, VENT), water demand for all AMs which used adiabatic cooling by evaporation, and labour costs for maintenance and service. Some of these cost components allow for synergies with climate change mitigation, such as photovoltaic investments on farm buildings, which have become mainstream for energy intensive livestock farms in some regions. The decision on whether a certain AM is economically beneficial can be appraised by a net present value calculation [18]. Decisive parameters typically are the assumed discount rate and future gross margins. Here we assume that gross margins are stable over time. Declining gross margins and a large discount rate would reduce the ER estimates of AMs and vice versa. Hence, sensitivity analysis could be a useful extension to increase the value of our results for farm decision makers. Decisions on AMs as presented in this study either fall into the range of incremental or systemic adaptations. Mitter et al. [41] highlight the long-term time horizon of both the decision making towards systemic adaptations and the duration of the invested fixed farm capital. It is another argument in favour of support for long term decision making.
The profitability is one aspect with typically high priority for farmers. However, the application of AMs multiple positive and negative effects. For example, impacts the welfare of animals, which becomes increasingly important for the European society. Several diseases are related to HS as a disturbance of the gut microbiome [42], an impacted intestinal barrier ("leaky gut") and high endotoxin concentrations in the blood [43], which might lead to local inflammations [44]. The improvement of the thermal environment and the air quality is a major aspect to ensure the five freedoms for animals and the concept of "life worth living" [45]. Therefore, reduced stocking densities, i.e. two of the modelled AMs, can serve two pressing societal needs, although with limited effectiveness on HS. A potential challenge for the effectiveness of some of the analyzed AMs could be recent trends towards outdoor ranges of confined livestock buildings, which are reasonable investment options for farmers today. Results on the outdoor heat stress index in this study support such decisions. Water availability under heat stress conditions is a further critical issue to be addressed in the future. Swine have a higher requirement for water with increasing temperatures and additional water is needed for several AMs [46,47]. Due to the fact, that water is a limited resource in many regions, water saving technologies are urgently needed [48]. The described AMs have to be assessed for their water saving potential in future studies. Fernandes et al. [49] reviewed costs and benefits for improving farm animal welfare highlighting that many improvements are difficult to evaluate from a purely economic perspective.
Several environmental side-effects are known for some of the investigated AMs. For AMs with a heat exchanger (CPHE, GEO) or a heating system that uses the inlet air below ground (EAHE), the ventilation rate can be increased during wintertime, which increases the air quality as well. This can be quantified by the indoor CO2 concentration. For days with a high diurnal variation of the outdoor temperature, the cooling and/or heating of the inlet air can damp this variation. Indoor air temperature has an effect on the release of ammonia, odorous substances and dust [50][51][52]. It can be expected that the application of AMs to reduce HS will also reduce these emissions. This is a relevant benefit, as for Europe about 94% of ammonia emissions are related to agriculture [53]. Livestock husbandry accounts for about 30 to 40%, besides manure storage and application and the use of fertilizers [54]. For Austria, a reduction of the ammonia emissions of about 26% has to be achieved on the basis of the national inventory of 2020 to implement the directives of the EU [55].
A drawback of the investigation presented here is the fact that only one single AM was simulated [56]. The effects from a combination of several AMs cannot be assessed on the basis of these data. The overall efficacy of a combination of AMs could be calculated by an additive or multiplicative operation, and the entire efficacy could show an augmentation or diminution by the combined use of AMs. The simulation of several AMs would be helpful to optimise the design values of such systems. Many reasonable combinations could be determined. All building AMs and some management AMs (e.g. INV, VENT, BREEDS, and FEED) could be combined with the air treatment AMs or the animal level AMs, but also building AMs and management AMs could be useful combinations with lower investment costs. On the other hand, the combination of the air treatment AMs and the animal level AMs does not seem to be useful.

Conclusions
We used a three-step process in order to assess the climate change related ER in growing-fattening pigs and the potential of risk reduction with AMs.
The AMs which are part of the ventilation system and cool the inlet air allow to substantially lower the degree of HS that can be expected inside growing-fattening pig buildings for the time horizon 2030. These measures reduce the occurrence of HS even below the 1980 level. Despite also influencing the occurrence and severity of HS, other AMs were shown to be less effective. For the reference scenario, the ER is estimated to be 4.98 € per year and animal place in 2030, with the VaR being 6.80 and 7.90 € a -1 per animal place for a 10 and 20 year return period, respectively.
Due to the fact that the ammonia emissions for livestock farms have to be reduced in the future, emission abatement by AMs that are cooling the livestock building is an additional benefit of large societal importance.
In future studies, the combination of different adaptation approaches should be considered in order to derive information on an optimum design of multiple AMs. The combination of building-related measures with those involving management options and treatment of the inlet air seems to be particularly promising.
The ranking of AMs based on their effectiveness to reduce HS is a decisive step towards AM decision support for farmers. Subsequent research shall estimate the operating and investment costs of AMs as well as systematically assess the costs and benefits beyond climate change adaptation. It can also inform the design of efficient policies to support farm investments such as included in Austria's strategic plan for the upcoming period of the EU Common Agricultural Policy (CAP). AMs require further innovations with respect to their effectiveness, resource intensity, costs, etc. The results in this paper can support farm-tech companies to determine the most promising fields of innovation.