Does Precision Technologies Adoption Contribute to the Economic and Agri-Environmental Sustainability of Mediterranean Wheat Production? An Italian Case Study

Abstract

The European Green Deal has set a concrete strategic plan to increase farm sustainability. At the same time, the current global challenges, due to climate change and fuels and commodity market crises, combined with the COVID-19 pandemic and the ongoing war in Ukraine, affect the need for quality food and necessitate the reduction of negative external effects of agricultural production, with fair remuneration for the farmers. In response, precision agriculture has great potential to contribute to sustainable development. Precision agriculture is a farming management system that provides a holistic approach to managing the spatial and temporal crop and soil variability within a field to improve the farm’s performance and sustainability. However, farmers are still hesitant to adopt it. On these premises, the study aims to evaluate the impacts of precision agriculture technologies on farm economic, agronomic, and environmental management by farmers adopting (or not) these technologies, using the case study method. In detail, the work focuses on the period 2014–2022 for two farms that cultivate durum wheat in central Italy. The results suggest that the implementation of precision technologies can guarantee economic and agri-environmental efficiency. The results could serve as a basis for developing a program to start training in farms as well as to suggest policy strategies.

1. Introduction

The transition towards a sustainable agricultural system is a priority to ensure the Sustainable Development Goals (in particular, SDGs 2.3 and 12.4) of the United Nations Agenda 2030, as well as the European Green Deal objectives. In particular, the European Commission has set a concrete strategic plan to reduce the use of chemicals and fertilizers, enhance biodiversity, and assist farmers in decision-making processes to increase farm sustainability. In addition, the current historical period and geopolitical framework lead to significant impacts on the agricultural sector. In particular, wheat production is currently affected by a significant stock depletion and price volatility. Starting with COVID-19 in 2020, the unexpected spread of the pandemic and the resulting lockdowns and closures around the globe led to an unavoidable critical situation related to the export restrictions and the changes to the purchasing behavior of wheat derivatives, such as flour [1]. These circumstances have put Europe and countries such as Italy in severe deficit conditions in terms of stocks, which also derives from the increased price volatility. Price volatility can be partially traced to uncertainty over the flow of supplies, depending principally on current production and existing stocks. The U.S. Department of Agriculture estimates that global wheat ending stocks for the 2022/2023 marketing year will be around 267 million metric tons. More than half of these stocks will be held by China, while EU, USA, and other major exporters account only for 20%. China’s wheat stocks increased by over 160% between 2012 and 2020. This was largely due to changes in China’s agricultural policy, which increased producer support prices, resulting in the accumulation of large government stockpiles [2]. By contrast, wheat stocks held by the rest of the world declined by 12% over same period.

Moreover, the ongoing war in Ukraine has contributed to reduce the wheat production in the country, disrupting the markets worldwide. The Russia–Ukraine war has caused the highest increase, since 2008, in levels and volatility of prices in agricultural markets for wheat, creating an ongoing vulnerability for global food security [3,4,5,6]. One difference between the two periods is the scale of the disruptions in staple food markets. While the period of initial pandemic lockdowns saw some isolated volatility, the Russia–Ukraine war is affecting all major food staples [7]. The relative tightness of global stocks suggests that price volatility will continue to remain high in respect to the past 10 years. Going forward, rebuilding inventories of wheat and other key global crops would help to reduce both prices and price volatility. By the same token, tight stocks mean that an unforeseen production shortfall in a major wheat producing region would likely send prices sharply higher again (as in 2010/11 and 2012/13) and result in increased price volatility.

In addition, fertilizer prices are a determinant factor more now than at the beginning of the pandemic, where the situation was already compromised. Even if the prices were at extremely high levels before the war began, they are still continuously rising; nonetheless, Russia, an important fertilizer producer, is considering an export ban. Furthermore, the energy crisis due to the high prices for natural gas, an essential feedstock to produce nitrogen-based fertilizers such as urea and ammonia, is contributing to boost fertilizer prices as well [8]. Higher fertilizer prices could depress production, leading to less grain on the market in 2022 and putting further upward pressure on already-high food prices [9,10]. In this context, it is important to analyze the cereal sector with reference to wheat production, which remains a mainstay of nutrition both in Italy and worldwide. This is because it is essential to understand how to cope with the current crises, considering the market dynamics that are being determined such as the rise in fertilizer prices and the volatility of wheat prices, factors that would make the cultivation of wheat (and cereals in general) unprofitable.

As a consequence of the global instability, the implementation of sustainable resilient strategies in agriculture is crucial. This entails the implementation of innovative agricultural practices that increase the productivity and income of farmers and, at the same time, can help to maintain ecosystems. Therefore, one of the actions to implement is reducing the quantity of inputs, in particular fertilizers, while maintaining production to protect both the environment and the income of farmers [11,12].

A key factor for sustainable agriculture is the introduction of digital technologies, which can help farm management through better-informed and timely decisions. These new technologies are known by the term Precision Agriculture Technologies (PATs), a farming management concept based on observing, measuring, and responding to inter and intra-field variability in crops [13]. According to an official report jointly published by ITU and FAO in 2020 [14], “digital agriculture has the potential to contribute to a more economically, environmentally, and socially sustainable agriculture, while meeting the agricultural goals of a country more effectively”.

From the beginning of the 1990s, different authors have discussed the agri-environmental and economic effects derived from the application of PATs [11,15,16,17,18,19,20,21]. In detail, most papers deal with environmental sustainability. The environmental benefits of precision agriculture (PA) derive mainly from the optimization of the management of crop inputs, such as as seeds, fertilizers (especially the efficient use of nitrogen), pesticides, irrigation water, and diesel, which often results in a reduction in their consumption without a decrease in the yield. It is notable that some studies report that the quantity of inputs does not decrease, but their use is optimized to avoid waste and pollution [12,22]. In addition, it emerges that, from an environmental point of view, through PATs it is possible to improve the soil proprieties (sustainable nutrient management) and reduce greenhouse gas emissions [23,24,25]. Finally, the optimal management of weeds is underlined [26,27].

The research on precision agriculture applied to the cereal farming started later, in 1997. In this scenario, the increasing interest of academia in this topic is notable from Figure 1, with an exponential increase in the number of documents (articles and reviews) available per year. The highest-producing countries regarding PA adoption in cereal farming cultivation are the United States, China, and Australia, while Italy ranks just fifth.

Considering the growing population and the necessity of safe food, establishing methods to increase the yield of staple crops, such as wheat, without compromising the sustainable development of future generations, is a challenging task. The implementation of PATs, such as variable rate application systems, could improve productivity, providing support to both producers and consumers [28,29,30,31,32,33]. In the pool of available documents, only few studies assess the economic sustainability of PA application in the cereal sector [34,35,36,37]. The economic benefits involve a general reduction of production costs, especially due to the correct management of crop inputs (reduction of pesticides and nitrogen) and an increase in productivity of the farm. The major economic benefit is recorded in the decrease in labor costs and the cost saving of fuel. However, an increase in total costs due to the capital invested in technology is highlighted.

However, adoption of PA tools is still far behind expectations, in part due to limitations in quantifying and demonstrating its economic and environmental benefits, insufficient detailed knowledge on technological functions, small farms managed by older farmers, and the deficiency of an incentive system [13,38,39,40,41,42,43,44,45,46].

On these premises, this paper evaluates the impacts of PATs on farm economic, agronomic, and environmental management by farmers adopting (or not) these technologies, using the case study method proposed by Yin (2009). This study is part of the activities of the Operational Group SMART AGRICULTURE TEAM financed by the Rural Development Program (RDP) Marche 2014/2020, sub-measure 16.1 (Appendix A, Figure A1). The objective of the project is to evaluate how PATs could support the optimization of nitrogen fertilizer management in durum wheat production. Contextually, the Operational Group aimed to evaluate the economic and environmental sustainability of cereal farms adopting or not adopting PATs. The work focuses on the period 2014-2022 for two farms (A and B) that cultivate durum wheat in central Italy. Farm A has used PATs since 2018; farm B uses conventional agronomic management. Based on the objective of the Operation Group, this paper tries to answer the following research questions:

• How does the durum wheat profitability evolve if a farm adopts or does not adopt precision agriculture technologies?
• Could the application of precision agriculture technologies improve and make more efficient the nitrogen use within the context under investigation?

The economic trend of durum wheat production is explored using a profitability ratio analysis. In addition, to understand what will happen to farm B if it decides to adopt the PATs package of farm A, a simulation was performed for the year 2022.

From an agri-environmental perspective, fertilization management is one of the most relevant targets of the PA. In particular, the nitrogen (N) derived from fertilizers, when inefficiently used in crop production systems, can move from agricultural fields and contaminate surfaces and groundwater resources, as well as contribute to greenhouse gas emissions (GHG) [47]. Since the interaction between the N rate, soil, weather, and crop response is a complex system, the management of this nutrient is the key aspect that distinguishes PA from conventional management [48,49]. Thus, the N environmental and agronomic efficiency is measured in this paper with the estimation of the nitrogen agronomic efficiency (NAE) index. This paper is structured as follows: Section 2 describes materials and methods; Section 3 presents the results and discussion. Finally, Section 4 concludes.

2. Materials and Methods

2.1. Study Area and Data Set

This work focuses on the period 2014–2022 (9 years) for two farms (A and B) that cultivate durum wheat (Triticum turgidum subsp. Durum Desf) in the Marche Region (central Italy) in rotation with maize (Zea mays L.) (Figure 2).

The climate of the study area is meso-Mediterranean based on the Walter and Leith Climate Class (Figure 3), which is characterized by a mean annual precipitation of about 768 mm and a mean annual temperature of 17.2 °C with monthly means ranging from 9 °C in February to 29 °C in August. There is a potential for frost from February until March and a period with a high probability of drought from June to August.

The physical and chemical compositions of the soil for the compared farms are reported in

Table 1. Soil physical and chemical compositions for the experimental farms.

Farm	Sand	Silt	Clay	pH	Organic Matter	Total Nitrogen
A	224.00	450.00	326.00	8.70	15.30	1.30
B	142.00	462.00	396.00	8.13	14.50	1.07

.

The two farms are agronomically managed differently; farm A acquired the first PA package in 2018 and started to adopt it in 2019. This period is considered the years of “technical change”, in which farm A has fully implemented the use of the PAT package considered in the present study. In line with the subdivision made by Finco et al. [21], the PA package acquired by farm A includes:

• Guidance systems (driver assistance, machine guidance, controlled traffic farming)
• Recording technologies (soil mapping, soil moisture mapping, canopy mapping, yield mapping)
• Reacting technologies (variable-rate irrigation and weeding and variable rate application of seeds, fertilizers, and pesticides).

In detail, farm A invested EUR 531,000 in PATs (Appendix A,

Table A1. Costs of the investments in the PA package for farm A.

Equipment	Type of Technology	Price (EUR)
AGCO challenger ISOBUS-ready tractor	Guidance system	200,000
Mazzotti MAF 5400 self-propelled sprayer with GPS guidance system	Guidance system and reacting technology	200,000
Seletron system	Variable distribution of pesticides	20,000
Jhon Deere ISOBUS-ready no till seeder	Guidance system	67,000
Sulki ISOBUS-ready fertilizer sprayer	Reacting technology	24,000
Parrot drone + software	Recording technology	5000 (+150 annual subscription)
Trimble in-cab terminal and satellite receiver combination for GPS guidance system and variable and rate distribution system Vantage Trimble	Guidance system and reacting technology	15,000 (+150 annual subscription)

). The investments in agricultural machinery equipped with PA technologies were financed for 40% of the total amount by joining Measure 4.1 (“Support for investments in farms”) of the PSR Marche 2014–2020. The use of this equipment is not limited to wheat cultivation; they are also used for the management of other cultivation, such as corn, on a surface that is four times larger than the one of farm B. In 2018, based on estimated cash flows at the time, the expected payback period (PBP) for the entire technology package was 5 years. Despite this, given the peculiar market trend during the period 2020–2022, the PBP dropped to 3 years, and currently the entire investment is paid off.

On the other hand, farm B used conventional agronomic management in all the years of this study. Figure 4 represents the experimental design of the case study.

As already mentioned, this study is based on an Italian EIP-AGRI Operational Group Project called the Smart Agriculture Team (SAT), whose main goals have been the following:

• To ensure a correct management of nitrogenous inputs on durum wheat through precision agriculture technologies in order to reduce the environmental impact of cereal cropping systems
• To evaluate the economic, environmental, and social sustainability of investments in these technologies.

Farm A was selected as a case study as it represents one of the few pioneering Italian farms that decided to adopt PATs. Likewise, farm B has been selected as a “control” case for not yet adopting PA technology after an in-depth analysis of all the Marche region farms associated with the largest trade association of Italian farmers. Thus, the three criteria on which farm B was examined were the following:

• The presence of a strong and real willingness to adopt the PA technologies investigated in this study
• The presence of a comparable size of the Utilized Agricultural Area (UAA) devoted to cereal farming and of a minimum total farm size of 100 ha to be defined as a large farm according to FADN statistical standards
• Farm B is a more efficient farm than the average in terms of productivity and profitability even without the implementation of PA technologies. In this regard, as it can be seen from the data (

  Table 2. Performance indicator comparison between FADN database and the two selected case studies. Index base value: central Italy.

  	UAA Durum Wheat	Average Yield	Average Yield Index	Average Durum Wheat Price	Average Durum Wheat Price Index	Gross Profit	Gross Profit Index	Operating Profit	Operating Profit Index
  	ha	t/ha		€/t		€/ha		€/ha
  Central Italy farms > 40 ha (2010–2018)	15.2	4.7	1.00	222	1.00	478	1.00	295	1.00
  Farm A (2014–2018)	87.6	5.66	1.18	257	1.16	776	1.62	498	1.69
  Farm B (2014–2018)	54.2	6.2	1.31	213	0.96	494	1.03	294	0.99
  Farm A (2019–2022)	103.0	5.4	1.14	392	1.76	1401	2.93	1019	3.45
  Farm B (2018–2022)	57.0	5.9	1.26	369	1.66	1275	2.67	1075	3.64

  ) that farm B is capable of levels of profitability almost in line to the median operating profit per hectare (calculated net of European CAP supporting payments applied to the durum wheat production) obtained by farms larger than 100 ha and specialized in cereal farming in central Italy. Furthermore, farm A and farm B are both located in the top 25%—in terms of operating profit per hectare from durum wheat farming—cereal farms in central Italy.

Farm A and farm B, while both producing durum wheat, are different from each other both structurally and from the point of view of the entrepreneurial and management logic that guides their strategic and operational choices. Farm A is a farm of about 400 ha, is cultivated using minimum tillage regime, and is almost entirely irrigated. Three quarters of the hectares are positioned in flat areas, and the rest are in hilly areas. Farm A’s mission is explicitly oriented towards technological innovation. About half of the farm UAA is used for the cultivation of cereals, including corn, while about a quarter of the UAA is used in the production of industrial legumes. It is important to note that farm A is integrated up-stream along the supply chain with an important Italian seed industry. Farm B is a farm of about 110 non-irrigated hectares cultivated using minimum tillage regime, located in hilly areas, and almost entirely occupied by cereal and forage crops.

For contextualizing the two case studies in the territorial framework, a comparison between them and our elaboration on the FADN sample of cereal farming in central Italy was carried out using basic profitability indicators, i.e., productivity, average value (price), gross profit, operating profit (Table 2). The historical series analyzed in Table 2 is divided into two periods according to the year of adoption of the PA by farm A in 2018.

Based on Table 2:

• Productivity: For both periods considered (2014–2018 and 2019–2022), the two case studies are both considerably more productive than the median value of productivity referred to in the sample of farms (greater than 40 ha) producing durum wheat in central Italy. Nevertheless, in the period 2019–2022, that is, the period after the acquisition of the PA technology by farm A, both farms A and B slightly lost productivity compared to their levels in the previous period.
• Price of the durum wheat produced: in the period 2014–2018, farm A proves to possess a capacity to enhance production with a notable premium price compared to central Italy (+16%) and farm B (+20%). This difference in price is due to the fact that farm A markets its product as seed wheat, a niche market in respect to the mainstream production of semolina wheat. In the period 2019–2022, post-PA adoption by farm A, the world changed drastically due to the double crisis (pandemic and the war in Ukraine) which, as we know, has led to a shock on the commodity market. Therefore, the surge in profit margins per hectare experienced by both case studies is due to the short-term economic prospects.
• Profitability (2014–2018): in the period 2014–2018, the operating income generated by every hectare of durum wheat produced by farm A was 69% higher than that of central Italy and 70% higher than that of farm B. This evidence indicates a much greater cost efficiency experienced by farm A in its PA pre-adoption period with respect both to the median context and to farm B. On the other hand, during 2019–2022, both case studies show an operating income which increased considerably because of the supply shock within the European market. In this regard, it is interesting to note that the difference in competitiveness between the two case studies observed in the previous period disappeared, as indicated by the operating income settling on the same level for both the farms.

The economic results obtained by farm A to produce durum wheat in the pre-adoption period (2014–2018) in comparison to that of the reference context are not surprising; in fact, from a managerial point of view, farm A is a farm characterized for being explicitly oriented towards efficiency and for having a very high propensity to innovate, which is an atypical attribute in the agricultural context investigated. As confirmed by the Smart AgriFood Observatory in 2021, the Italian UAA managed with precision agriculture techniques is around 4%.

Farm A relies on a managerial structure given by three managers—i.e., the managerial structure coincides with the farm ownership—plus three full-time workers (all three highly skilled agricultural technicians). One of the three managers is a young, specialized technician who has been responsible for the computerized and automated farm management since the acquisition of the PA technology in 2018.

Despite being a larger and more profitable wheat producer compared to the median value of the sample of cereal farms in central Italy, farm B is characterized by a traditional management structure which does not employ full-time workers and where the management work and the work in the fields are both carried out directly by the entrepreneur and his family.

Finally, focusing on nitrogen management,

Table 3. Agronomic management practices of the farms.

Farm A
Field Activities	Period
Ploughing (40 cm)	October
Harrowing	November
Sowing	November
Pest control: Azoxystrobin, Cyproconazole	March
1st N fertilization—VRT 1	March
2nd N fertilization—VRT 1	April
Harvest	July
Farm B
Field Activities	Period
Chisel (25 cm)	October
Harrowing	November
Sowing	November
Pest control: Azoxystrobin, Cyproconazole	March
1st N fertilization	March
2nd N fertilization	April
Harvest	July

¹ VRT: Nitrogen fertilization performed with the Variable Rate Technology.

lists all the practices applied by both farmers, acquired through the field notebooks.

2.2. Economic Analysis

The economic analysis aims to explore farm profitability in adopting or not adopting PATs through indicators by comparing two case studies (farm A and B) based on Yin’s case study design [50]. This approach was chosen because the focus of the study is a contemporary phenomenon characterized by a small number of pioneers adopting PA.

To carry out this study, a profitability analysis was performed employing financial ratios [51,52]. This analysis allows comparing the two cereal farms, A and B, placed in the same locality and similar in the UAA devoted to the production of durum wheat; in this way, the understanding of the discrepancies in the results, determined by a different management approach and in the propensity to adopt new technologies, can emerge [53]. We restate that farm A has invested in precision farming technology since 2018, while farm B has not (yet) invested in precision farming technology but operates under the conventional management system, and it is considered a possible “target farm” that could adopt PA.

Thus, we designed our case study as follow:

• The profitability of durum wheat production performed by the PA-adopting case study (farm A) has been assessed by comparing how the profitability indicators evolve before and after the adoption period (2014–2018 vs. 2019–2022).
• Besides, the profitability of durum wheat production has been assessed comparing the economic indicators of the PA-adopting farm (farm A) to that of the non-adopting farm (farm B).

By being limited to a specific crop, the analysis has been conducted using margin ratios (income statement analysis) as indicators of profitability but not return ration (balance sheet analysis), since this type of indicator would have required an analysis of the profitability of the farm business taken as a whole. Instead, this study focuses only on durum wheat profitability, meeting the objectives of the Operational Group SMART AGRICULTURE TEAM financed by the Rural Development Program (RDP) Marche 2014/2020, sub-measure 16.1.

It is also important to point out that this economic analysis was not constructed as an experimental field trial but as a comparative case study conducted within real farms operating on the real market. In fact, our goal is not to directly (experimentally) evaluate the effect of some PA device on the crop profitability; rather, the objective is to analyze basic crop profitability measures and indices during the period of the PA adoption process. In this regard, while supporting the necessity of carrying out experimental trials to verify the economic efficacy of adopting specific technologies to specific crops, we underline that also the economic effectiveness evaluation of technology adoption carried out in the “real farm” productive space can generate further elements of analysis useful in understanding the determinants of the adoption process. Our work falls into this second category of studies on technology adoption effectiveness.

The measurement and ratios [52] utilized to perform the profitability analysis refer to durum wheat production, and are listed below:

• Productivity

  ○

  T/ha

• Gross Profit (per hectare)

  ○

  Revenues (RV) − Variable Costs (VC)

• Gross profit Margin

  ○

  (RV − VC)/RV

• Operating profit (per hectare)

  ○

  Gross Profit − (PA capital depreciation quota − land for rent quota − administrative and general expenses quota)

• Operating profit margin

  ○

  Operating profit/RV

All the data useful for this analysis were obtained by means of in-depth interviews of the agribusiness entrepreneurs of the two farms.

2.3. The Nitrogen Agronomic Efficiency Index (NAE)

To measure the environmental and agronomic efficiency, the nitrogen agronomic efficiency (NAE) index was calculated by the following formula (Equation (1)):

$$\mathrm{N}\mathrm{A}\mathrm{E}=\frac{\mathrm{Y}\mathrm{i}\mathrm{e}\mathrm{l}\mathrm{d} \mathrm{h}\mathrm{a}\mathrm{r}\mathrm{v}\mathrm{e}\mathrm{s}\mathrm{t}\mathrm{e}\mathrm{d} \left(\mathrm{k}\mathrm{g}/\mathrm{h}\mathrm{a}\right)}{\mathrm{N}\mathrm{i}\mathrm{t}\mathrm{r}\mathrm{o}\mathrm{g}\mathrm{e}\mathrm{n} \mathrm{p}\mathrm{r}\mathrm{o}\mathrm{v}\mathrm{i}\mathrm{d}\mathrm{e}\mathrm{d} \mathrm{t}\mathrm{o} \mathrm{t}\mathrm{h}\mathrm{e} \mathrm{c}\mathrm{r}\mathrm{o}\mathrm{p} \left(\mathrm{k}\mathrm{g}/\mathrm{h}\mathrm{a}\right)}$$

(1)

The NAE is the ratio between the total yield harvested (kg/ha) and the nitrogen provided to crop (kg/ha). The higher the NAE value, the greater the nitrogen use efficiency for production purposes. At crop maturity, the yield data was collected with a combined harvester for the entire durum wheat production area. The yield data (t/ha) was calculated from measurements taken at the time of delivery to the consortium.

3. Results and Discussion

3.1. Economic Results

In this paragraph, the main economic results will be presented.

Table 4. Economic analysis.

Harvest Year	Productivity (t/ha)		Durum Wheat Price (€/t)		Variable Costs (€/ha)		Variable Cost Ratio (A/B)	Gross Profit (€/ha)		Gross Profit Ratio (A/B)	Operating Profit (€/ha)
	A	B	A	B	A	B		A	B		A	B
2014	6.40	6.20	287.30	230.00	580.40	702.50	0.83	1183.32	598.50	1.98	889.89	398.50
2015	4.90	5.80	306.40	240.00	600.50	712.50	0.84	817.86	554.50	1.47	511.60	354.50
2016	5.70	5.80	213.00	180.00	588.80	672.50	0.88	576.30	246.50	2.34	409.28	46.50
2017	5.60	6.00	245.00	200.00	564.85	677.50	0.83	753.84	397.50	1.90	377.41	197.50
2018	5.20	7.00	234.00	215.00	600.55	709.50	0.85	548.99	670.50	0.82	300.90	470.50
2019	5.60	5.50	270.00	245.00	572.80	662.50	0.86	898.87	560.00	1.61	410.22	360.00
2020	5.90	6.50	326.60	270.00	592.58	698.50	0.85	1296.09	93.50	1.39	965.21	731.50
2021	5.30	5.80	480.00	470.00	605.00	692.50	0.87	1879.00	1908.50	0.98	1528.29	1708.50
2022	4.70	5.50	490.00	490.00	771.40	1017.50	0.76	1531.60	1699.50	0.90	1170.97	1499.50

shows the comparison between farm A and B from 2014 to 2022 in terms of productivity, cost efficiency of the production process, and profitability.

(1)

Productivity: Land productivity is a very complex indicator that depends on many variables involved. In our case study, the data show that the most productive farm is the one that does not adopt the PA: farm B. Moreover, what is noted is also a slight declining trend in productivity for both farms, and perhaps this evidence could be related to the change in atmospheric and climatic conditions in the medium term. However, this is a hypothesis that should be verified using statistically representative samples of cultivated areas. Then, focusing the attention on the post adoption period, we note that farm A shows an increase in productivity in the 2019–2020 period followed by a decrease in productivity in the period 2021–2022. Again, the owners/managers of farm A attribute these trends as essentially linked to environmental conditions and not directly linked to the use of PATs which, among other things, should not be a factor of productivity increase but of cost optimization for any given level of productivity.

(2)

Cost efficiency: Regardless of the use of the PATs, looking at the trend of variable costs and the variable costs ratio, it emerges that farm A is a farm structurally more efficient than farm B, while, in terms of PA cost effectiveness, until 2021, the variable costs ratio remains substantially constant. Therefore, no signs of PA adoption efficacy are observed. Things change in 2022. Indeed, the variable costs ratio between farm A and farm B falls from 0.83–0.87 (in trend) to 0.76. Although this is an observation of only one year, so not very meaningful if seen in isolation, it still allows us to make a hypothesis: with raw material prices at the levels of 2022, the cost optimization of the production process using PATs could become significant and relevant. Obviously, this hypothesis should be tested experimentally; nevertheless, our data indicate that the farm that adopts a PAT management shows resilience in terms of increase in the production cost per hectare, which is much greater than the case study that does not adopt PAT.

(3)

Gross profitability: Interesting information can emerge if observing the gross profit. First, in the pre-adoption period, farm A was shown to be capable of much higher profitability than the “control” case study (farm B). Since 2018, in conjunction with the investment in the PAT package, farm A apparently loses its profitability advantage with respect to farm B. Indeed, in the period 2021–2022, the gross profit ratio between the two case studies is reversed compared to previous years—the wheat produced by farm B becomes more profitable than that produced by farm A—and this is due to three underlying forces acting simultaneously: wheat selling price, productivity, and contingency of exceptional environmental conditions.

• Selling price: Since 2018, the difference between the two case studies in terms of average revenue narrowed, until it disappeared in 2022. The exceptional increase in prices in the three-year period, 2020–2022, favored an upward squeezing of the price differentials, which was previously linked mainly to product quality.
• Productivity: Farm B remains a structurally more productive farm even in the post-adoption period of the PA package by farm A. The higher productivity of the durum wheat produced by farm B lies in the genetics of the seeds used. Farm A produces durum wheat for seed. The varieties used are generally less productive than semolina varieties, but they usually tend to have a higher market value even if, as we have seen in 2021–2022, the price of the two case studies flattens out on the same level due to the market shock.
• Environmental conditions: Although the use of PATs allows a greater timeliness of action in crop management, even without the use the technologies, farm B was able to manage the 2021 sowing period more effectively than farm A. The 2021 sowing was very difficult in the survey area due to exceptionally prolonged rain events. Farm A was unable to sow before December 2021 (two months of delay), and this strongly influenced the low productivity of the 2022 harvest, while farm B found useful windows for sowing in the right period, i.e., October 2021.

(4)

Operating profit: the fundamental information contained in the comparison between the two case studies, in terms of operating profit, is the incidence of the depreciation share of the PA capital invested by farm A in 2018. This factor, combined with the alignment of the prices of wheat sold starting from 2020 and the higher productivity of farm B, determines an inversion of the profitability of the two case studies in 2021–2022, when farm B becomes more profitable than farm A. The weight of the share of depreciation of the PA capital on the profitability per hectare of farm A also emerges from the joint comparison of the gross margin and the operating margin (Figure 5). In fact, the narrowing of the distance between the two indicators that can be seen when passing from the gross margin to the net margin is essentially due to the depreciation rate of the PA capital discounted by farm A.

In 2020–2022, the operating profit of farm A improved to levels far above pre-adoption conditions, and this is especially due to the market price trend (Figure 6).

In addition, farm adopting PATs creates advantages over traditional farming in terms of better management of resource efficiency. This aspect is particularly relevant for the use of N fertilizer. In fact, after COVID pandemic and for the Russia–Ukraine war, this input increased its price by 176% from January 2020 to December 2022 (Figure 7).

In this scenario, PATs allowed farm A to optimize N distribution according to the specific necessity of the crop as shown in the following paragraph (NAE index).

In this way the farm A works achieving both a better quality of production and minimizing the negative impacts on the environment.

Finally, to understand what will happen to farm B if it decides to adopt the PATs package of farm A, a simulation was performed for the period 2020–2022 (

Table 5. Simulation of PAT adoption for farm B for the period 2020–2022.

Year	Average Yield (t/ha)	Durum Wheat Price (€/t)	Variable Costs (€/ha)	Operating Profit (€/ha)	Operating Margin
2020	6.50	270.00	698.50	444.77	0.25
2021	5.80	470.00	692.50	1416.56	0.52
2022	6.00	490.00	1017.50	1202.24	0.42

). A depreciation cost of the same PA capital acquired by farm A is considered with a variating depreciation rate according to the durum wheat farm UAA.

It emerges that, if farm B had acquired the same PA package as farm A, the operating margin of farm B improves thanks to the new market conditions in the period 2020–2022 despite the cost of PA capital. This evidence suggests that PA adoption by farm B could be feasible in economic terms thanks to a sufficiently profitable, productive, and extensive farm structure in which implementing the new technologies in this new market conditions (which, however, are constantly changing).

Nevertheless, despite the favorable economic situation, farm B is currently not prone to technological change. The motivation could not be purely economic, but it could be linked to the characteristics of the owner. As the literature suggests [54,55,56], older farmers show a lower propensity to adopt as compared to their younger counterparts. Old farmers’ may be loath to changes and they may not see longer-term benefits perhaps because they lack training and their bond to conventional agricultural management [57]. Moreover, access to credit is certainly another possible constraint to adoption.

3.2. Agronomic Results

Farm A supplied on average less nitrogen (−63%) than farm B for each year under analysis (

Table 6. Total nitrogen provided, crop yield, and NAE per farm each year.

Year	Farm	N Provided (kg N/ha)	Tot. Yield 1 (kg/ha)	NAE 2
2017	A	136	5600	0.41
2018	A	129	5200	0.40
2019	A	114	5600	0.49
2020	A	177	5900	0.33
2021	A	125	5300	0.42
Mean A		136	5520	0.41
2017	B	210	6000	0.29
2018	B	230	7000	0.30
2019	B	215	5500	0.26
2020	B	223	6500	0.29
2021	B	208	5800	0.28
Mean B		217	6160	0.28
Mean (A vs. B) differences (%)		−63	−10	+47

¹ The yield production (kg/ha) referred to a humidity content of 13%. ² NAE: Nitrogen Agronomic Efficiency.

). While evaluating the average yield, during the five growing seasons, it shows that farm B achieves 10 percent more than the farm A.

The NAE index, which is an index designed to assess nitrogen fertilizer use efficiency, shows that farm A obtained a higher value (+0.47) than farm B (Table 6).

After water, nitrogen fertilization management is the most important plant nutrient [58]. Nitrogen fertilization contributes significantly to crop development, chlorophyll accumulation [59,60,61,62], and nitrogen content [63].

Mineral nitrogen contributes to better growth by giving the crop the nutrient when it needs it most, which results in higher production [59,64] and quality grain levels [65]. Farm B obtained a higher production level than farm A due to the higher nitrogen provided to the crop.

Ref. [66], with 20 years of data on durum wheat production, shows that nitrogen is the key driver of the production. The authors have shown that increasing the nitrogen supplied to durum wheat allows a significant increase in yield.

It is also true that as the dose increases, the yield of durum wheat does not increase proportionally. Ref. [67] showed that nitrogen doses above 150 kg N/ha do not increase yield but, on the contrary, result in a higher protein percentage.

When the nitrogen is not absorbed by the crop, it can only have two fates, leaching [68] and denitrification [69], which have negative environmental impacts, without considering the economic damage suffered by the farmer and that such production inputs are less available and increasingly expensive.

Given its consequences at the agronomic, economic, and environmental levels, the management of nitrogen fertilization has always been an important topic of scientific research [28]. Today, to optimize nitrogen fertilization at the farm level, precision farming has a strong impact on both the environment and the economy [63,70,71].

Precision agriculture is agronomic management based on the spatial and temporal variability of agronomic components, such as the soil’s chemical and physical variability [49] and crop needs. Analyzing spatial and temporal variability, prescription maps [62] [72] can be generated that allows the nitrogen dose to be adjusted according to crop needs and therefore improve the nitrogen use efficiency (NUE).

Several authors have reported that precision farming allows an increase in NUE. Ref. [73], in China, showed the yield and NUE results of precision agronomic management. The authors report an increase in yield and NUE compared to conventional agriculture of 10% and 51–97%, respectively.

In Switzerland with winter wheat (Triticum aestivum) [74], it was reported that precision nitrogen management improved the NUE by an average of 10%. Moreover, in Umbria, Italy, it was reported that the variable rate technology improved the NUE by 15% compared to the flat rate [75].

In accordance with all the previous works, an increase in the NUE was also achieved in our case study. Farm A, which uses the variable rate technology, obtained a higher NAE of 15% than farm B, which distributes nitrogen evenly. Farm A is more environmentally and profit-friendly than farm B.

4. Conclusions

This study is based on a double case study in central Italy and explores durum wheat profitability and the optimization of the nitrogen fertilization as a function of the management of the production process through PA technologies for the period 2014–2022.

Farm A acquired and implemented the PA package in 2018–2019, while farm B has not (yet) invested in PA technology but works under the conventional management system. Therefore, it can be considered a possible "target" farm that could adopt PA since it is a larger farm compared to the local farm average size. Moreover, the owner of farm B shows a high propensity to adopt these technologies and participated in this study precisely to have more points of reference for deciding on possible PA investments, especially in light of the growing cost of production inputs.

Since the adoption of PA is still at a pioneering state in central Italy, our case study can represent a useful benchmark for both agricultural entrepreneurs and policymakers with respect to the economic effects of PA technology adoption applied to durum wheat production. In detail, from the economic analysis, it emerges that, in terms of gross profit, there are substantial differences between the two case studies. Farm A is characterized by a gross profit that is, on average, higher than farm B in the pre-crisis period 2014–2020. Farm A’ s economic indicators have been affected by the PAT depreciation schedule coinciding with the technological change. Despite this, the economic efficiency of farm A improved to levels above pre-adoption conditions, thanks to the new market conditions in the period 2020–2022. In addition, farms adopting PATs optimize the use of inputs such as nitrogen fertilization according to crop needs; at the same time, it favors the farm management’s efficiency in terms of human resources.

In the 2014–2021 period, our study did not show any clear savings in terms of wheat production costs that could be attributed to the use of the PA technology package by farm A. However, things changed in 2022. In fact, with the surge in the price of inputs, the index of variable costs of farm A increased by 29%, while that of farm B increased by 46% with respect to 2021. The hypothesis is that this 17% difference, corresponding to about EUR 60 per hectare, could be due to the use of PATs by farm A. Despite this possible savings in terms of variable costs, it is necessary to consider that:

• The depreciation share of the financial capital invested by farm A in the PA package was EUR 89.18 per hectare in 2022.
• The agricultural area on which this share of depreciation is calculated is four times higher than the agricultural area available to farm B. As mentioned by Schimmelpfennig [76], large farms may present economies of scale when adopting PATs because they have more hectares over which to spread investment costs. Moreover, large farms are also more likely to have the type of variability that makes PATs [77].

Thus, because of the depreciation share, farm B, while being able to save around EUR 60 per hectare in terms of variable costs under the exceptional market conditions of 2022, does not show a broad economic incentive to invest financial capital in PA, probably due to a lack of economies of scale, while farm A does. These findings confirm our previous analysis relating to the dimensional thresholds necessary to create an economic incentive for investment in PA by a specialized cereal farm in Italy which is at least 200 ha in a hilly area [78]. Nevertheless, the incentive in using PA technology could be present even for smaller farms in terms of payment of a rent for a PA-type management assistance at service, rather than in the purchase of technological capital.

Regarding the willingness to adopt PA technologies for the cultivation of wheat, Hanson et al. [79] verified that, in North Dakota, wheat may have negative effects on PAT adoption due its lower cost of production with respect to other crops such as corn. This assertion is confirmed by our case study, given that corn is a fundamental crop within the farm A production structure while it is absent within the production structure of farm B.

This study is not without limitations. The first limitation of this study is the fact of having compared the pioneering case study A with a single control farm (B) rather than with a pool of farms. There were two constraints which prevented the initial intention of comparing case study A with a sample of farms:

• The research project from which this article derives involved an agronomic experimentation in the case study farms, and it would have been beyond the possibilities of the project to develop this experimentation in more than two farms (farm A and B).
• Therefore, the first issue was to identify a “control” farm (case study B) available to host the agronomic experimentation for the participation in the comparative study. This farm should have been available to provide all its economic and accounting data. In this regard, it should also be clarified that farms in Italy in many cases do not keep detailed analytical accounting relating to long historical series in their archives. For this reason, even working with just one farm, it was not easy to reconstruct the analytical accounting data set necessary for carrying out this study [80]. On the other hand, there were no difficulties with farm A, since it keeps track of its own detailed analytical accounting using advanced management software (as Geofolia, Isagri).

The second limitation consists in having only one farm (farm A) implementing a broad package of PATs and managed according to a logic that we can define as PA-oriented. However, this limitation could be explained by the fact that there is a lack of diffusion of these technologies, so it becomes rather impossible to work with broader samples, and it becomes necessary to develop research based on a few case studies. Therefore, according to this second limitation, it would be appropriate to create an infrastructure that allows researchers to be able to acquire reliable economic, financial, and agronomic data on which to perform analyses on the effects of PA technologies. In addition, being a pioneering technology in this area, few farms have purchased and adopted precision farming techniques because they would like to understand whether such technology provides an economic and environmental benefit. This aspect is closely linked to the objective of the present work. In conclusion, policymakers are advised to encourage the adoption of these technologies given that the current market conditions generate incentives to adopt, specifically, very high costs of input but very high prices of output [81]. Finally, both from an economic and an agronomic point of view, it is important to consider these aspects in order to appreciate all the advantages of this type of innovation that hinges on the automation of the production process. The farm deciding to adopt PATs must already possess both the characteristics and the philosophy of efficiency, since adopting promising and efficient technologies on obsolete or inefficient production systems—similar to what happens in the field of automation of manufacturing production processes—does not mean innovating but automating the pre-existing inefficient production processes. In addition, the adoption of PA technology requires training programs for farmers and farmworkers in order for them to acquire the right skills. In fact, the availability of training is a condition necessary for understanding and mastering the PA package characteristics, as well as for fully exploiting its potential in terms of efficiency and effectiveness.