Economic Effects of Sustainable Weed Management Against Broomrape Parasitism in Industrial Tomato

Abstract

Sustainable Weed Management Practices (SWMPs) are currently underrepresented in European cropping systems despite considerable attention from the research and policymaking communities. In public discourse, their adoption is associated with low yields, high initial investment costs, additional machinery requirements, elevated labor demands and limited or uncertain profitability. Nevertheless, little is known regarding their economic effects when implemented under real-life conditions at the farm level. This study aims to determine the impact of SWMPs against broomrape parasitism on the organization, management and economic performance of industrial tomato farms, considering that broomrapes (Orobanche and Phelipanche species) are a major impediment to the expansion of key crops in the Mediterranean basin due to their resistance toward commonly applied herbicides. For the purpose of economic appraisal, detailed technical and economic data were collected in 2022 from 76 arable farms cultivating industrial tomato in the Region of Thessaly in Central Greece. By combining Principal Component Analysis (PCA) with Two-Step Cluster Analysis (TSCA), a farm typology according to the implementation level of different SWMPs was developed. Based on this typology, a comparative technical and economic analysis revealed important differences in terms of structure, resource utilization and economic performance across the various farm types. “Holistic” farms, which exhibited the highest adoption levels of SWMPs, implemented an effective broomrape management strategy and achieved superior economic outcomes, evidenced by a remarkable net profit of 488.5 €/ha. Conversely, this was either negative or nearly negligible in farm types characterized by low adoption rates, indicating a lack of economic viability in the long run. The findings of this study offer useful recommendations for farm-level decision making, advisory support and policy design toward the promotion of SWMPs.

1. Introduction

Tomato (Solanum lycopersicum) is a prominent horticultural crop worldwide, consistently ranking top in vegetable production over the past thirty years [1]. Among its basic cultivation categories is industrial tomato, which holds a particular economic relevance for agriculture across the Mediterranean region owing to its high yields and processing potential [2]. In Greece, industrial tomato cultivation is widespread in many agricultural areas, serving as a vital source of income for rural households and providing long-term employment opportunities in local agro-industries. Despite its dynamics, industrial tomato is currently underrepresented in Greek agriculture. From around 19,000 ha cultivated in 2004, its acreage dropped to 4526 ha in 2021, representing less than 1.0% of the country’s total arable land [3]. This 76.0% decline over a 17-year period is mainly associated with the emergence of parasitic weeds commonly known as broomrapes (Orobanche and Phelipanche species), which adversely impact productivity and hinder the expansion of major crops in the Mediterranean basin, such as legumes, tomatoes, sunflower and oilseed rape [4]. Due to their distinctive biological and ecological characteristics (e.g., high seed yields, achlorophyllous nature implying reliance on host plants for nutrients, root-specialized parasitic growth, underground development, prolonged soil dormancy, high specificity for certain host plants) broomrapes cannot be classified as typical, non-parasitic weeds [4,5] and thus cannot be effectively controlled by conventional management strategies including chemical herbicide application [6]. Besides failing to contribute toward satisfactory control, chemical use in general raises concerns regarding environmental degradation and the development of herbicide-resistant weed species [7,8].

In this context, a plethora of non-chemical alternatives, involving, among others, cultural (e.g., crop rotation, no tillage, delayed sowing), physical (e.g., hand weeding, solarization), biological (e.g., insects, fungal pathogens), mechanical (e.g., mechanical weeding, deep plowing), host resistance (e.g., tolerant varieties, transgenic crops) and novel technological (e.g., precision agriculture tools) approaches have been extensively presented in the scientific literature throughout the years [4,9,10,11,12]. These alternatives are referred to in this study as “Sustainable Weed Management Practices” (SWMPs), which are intended, in conjunction with targeted herbicide use, to mitigate yield losses caused by broomrape parasitism without compromising the environmental and economic sustainability of farming systems. Relevant practices have been promoted through policy actions as a part of Integrated Weed Management (IWM) since 2009, following the implementation of the European Union’s (EU) Directive 2009/128/EC [13]. Currently, a number of SWMPs are included in the eco-schemes of the Common Agricultural Policy (CAP) which motivate farmers to implement practices that contribute to the EU’s Green Deal goals by providing them with supplementary income support. Therefore, the critical role that SWMPs can play in ensuring one of these major objectives, which calls for a 50% reduction in overall chemical pesticide use by 2030 [14], is widely acknowledged in the policy arena. In Greece, such practices in arable land comprise cover crops, precision technologies as well as mechanical weeding combined with mulching use, aiming to support biodiversity, agroecology and precision farming.

Although SWMPs have attracted considerable attention within the research and policymaking communities, their adoption remains limited in most European countries [15], including Greece [16], while broomrape parasitism persists. In public discourse, such practices have been connected with low yields [17], high initial investment costs [18], limited or uncertain profits [19], additional machinery requirements [20] as well as high labor demands [21], rendering them vulnerable to competition from chemical herbicides, which are perceived as a cheaper and less complex solution [22,23]. Even under a favorable regulatory environment, which is also reflected in the withdrawal of numerous herbicides from the EU marketplace [24], SWPMs are encountering difficulties in establishing their role within an IWM system. Nevertheless, while they have undergone sufficient assessment in terms of their agronomic efficacy [25,26,27], less interest is garnered on their relation to farm organization, management and profitability. Lacking empirical, clear evidence on the economic effects of SWMPs when implemented under real-life conditions at the farm level increases farmers’ risk and uncertainty related to the economic viability of such practices, which stands as a crucial factor in their decisions regarding the adoption of sustainable crop protection [15].

To date, there is a notable lack of scientific literature examining the impact of SWMPs on farm economic performance. Economic aspects of SWMPs, encompassing adoption costs, economic risks, benefits and financial returns have been thoroughly discussed mostly in review studies [8,28,29,30]. A substantial body of international literature has investigated the economic implications of such practices based on their technical performance in experimental settings, aiming at quantifying the trade-offs between environmental and economic sustainability [31,32,33,34]. However, none of these studies treated them as integral components of a cropping and economic system but rather as discrete technical interventions. The few existing farm-level impact research is exemplified in studies assessing particularly the effects of Integrated Pest Management (IPM) practices on farm economics [35,36,37,38,39]. Whereas most of these studies overlooked adoption intensity by categorizing sampled farms as either IPM-adopting or non-adopting [35,38,39], Midingoyi et al. [37] classified them into four distinct categories according to varying levels of IPM implementation and analyzed their income disparities. Mulungu et al. [36] also performed a comparative assessment to demonstrate how different combinations of IPM strategies translate into economic outcomes in mango fruit farms in Central Kenya. In any case, all these studies aimed to measure the partial effects of such practices on certain economic indicators (e.g., gross or net income) but did not deliver an integrated evaluation of farm overall economic performance reflecting their viability and long-term prospects. To our knowledge, the application of a typology to categorize farms under broomrape pressure based on the level of adoption of different SWMPs for the purpose of comparing their financial results through a detailed economic appraisal has not been previously documented in the literature.

Using a typology-based approach, this study aims to analyze the effects of SWMPs for broomrape control on the economic performance of farms cultivating industrial tomato in the Region of Thessaly in Central Greece. In particular, we employed data from a farm management survey to present the findings of a comparative technical and economic analysis based on a typology of 76 arable farms classified according to their implemented SWMPs. In addition to helping to understand heterogeneity through the identification of diverse farm profiles and their patterns, farm typologies are also crucial for developing tailored policies and strategies to facilitate the transition to sustainable agriculture [40].

This study contributes to the sparse current literature on the farm-level economic impact of sustainable crop protection in two ways. First, most of the existing studies measure adoption intensity as a binary decision, while very few consider the extent to which such practices are applied or explore the impact of their different combinations. We develop a farm typology by combining Principal Component Analysis (PCA) and Two-Step Cluster Analysis (TSCA) to examine how combinations of SWMPs, when implemented to varying degrees, influence the economic viability of industrial tomato farms in the broomrape-infested Mediterranean context. Second, we provide for the first time a comprehensive overview of the relation of SWMPs to farm structure, management and economic performance through an in-depth economic assessment. The results of this study could offer valuable insights for farm-level decision making, advisory support and policy design toward the promotion of SWMPs in Mediterranean and Greek agriculture.

2. Materials and Methods

2.1. The Study Area

The research area in this study comprised the Regional Units of Karditsa and Larissa, both of which are part of the Region of Thessaly in Central Greece. Thessaly is one of the main agricultural districts in the country with significant crop and livestock production. Its climate is Mediterranean with continental influence, as it is defined by cold and wet winters and hot and dry summers. The average annual temperature in the Region is about 16.0 °C. January is the coldest month, with a mean temperature of 5.2 °C, while July is the hottest, with a mean temperature of 27.2 °C [41]. The rainfall ranges significantly, from 360 mm per year in coastal regions to 1850 mm in mountainous parts, with an average annual precipitation estimated at roughly 780 mm [42]. Well-known for its fertile plain and vibrant rural heritage, Thessaly constitutes an agriculture-oriented economy, producing three times more output and added value in the primary sector compared to the national average [3]. Covering a total cultivated area of 347,069 ha (15.0% of Greece’s total cultivated area), the plain of Thessaly ranks among the largest agricultural plains of Greece and is dominated by annual arable crops (75.0% of Thessaly’s total cultivated area), primarily cotton and cereals [3]. Industrial tomato occupied around 2500 ha in 2021, corresponding to 1.0% of Thessaly’s total arable area. Nevertheless, this acreage accounts for 55.2% of the total area cultivated with industrial tomato in Greece, which renders Thessaly the flagship of industrial tomato production throughout the country. Farms cultivating industrial tomato are mainly located in the southern part Thessaly, particularly in rural communities of the Regional Units of Karditsa and Larissa (Figure 1).

In September 2023, the study area became the epicenter of Storm “Daniel”, a once-in-a-1000-year weather phenomenon characterized by extreme rainfall. Annual crops including cotton, maize and industrial tomato (i.e., the medium/late varieties) experienced total devastation due to the floods occurring just prior to their harvest period [43]. Following the severe repercussions of Storm “Daniel”, a strategic plan was developed [44] to recover agricultural production in Thessaly, focusing on crop restructuring among other objectives. Particularly, priority was given to crops with low input requirements as well as to more intensive crops with high market value such as industrial tomato.

Figure 1. Map of Greece with a focus on the Region of Thessaly [45,46]. The study area consists of rural communities in the Regional Units of Karditsa and Larissa.

Figure 1. Map of Greece with a focus on the Region of Thessaly [45,46]. The study area consists of rural communities in the Regional Units of Karditsa and Larissa.

2.2. Survey Profile and Data Collection

This study uses data from a farm management survey of farms cultivating industrial tomato, conducted from December 2021 to September 2022. Thus, it was completed approximately one year before the occurrence of Storm “Daniel”. A total of 76 farm heads participated in the survey, whose farms occupied 4824 ha of arable land (1.9% of Thessaly’s total arable area), 17.2% of which was cultivated with industrial tomato. This implies that the sampled farms accounted for 33.2% and 18.3% of the total area cultivated with industrial tomato in Thessaly and Greece, respectively. In addition, the sample reflected the broader operational and sociodemographic characteristics of farms across regions, considering the relatively homogeneous agroecological conditions prevailing in the industrial tomato sector across the country. In particular, they were indicative of farms operating within an intensive production system and specializing in arable farming, primarily in cereal and cotton production, with 15–20 years of engagement in the cultivation of industrial tomato. Nevertheless, they had encountered broomrape infestations over the past decade, which limited their ability to integrate industrial tomato into their cropping pattern annually. Available chemical solutions were initially employed to combat broomrapes in these farms; however, the appropriateness of this method was immediately called into question due to its detrimental effects on the environment and human health and its ineffectiveness against parasitism. In response to the pressing need for sustainable weed control, the surveyed farmers tended to experiment with alternative solutions. Therefore, they were representative of farmers somewhat familiar with SWMPs, as they had used some of these methods in their farms. Regarding sociodemographic attributes, they were typical of farmers who owned and managed their farm, had completed secondary education and had been involved in the farming profession for more than 20 years.

Data were gathered through in-person interviews with farmers using a carefully designed questionnaire. Potential participants were contacted at random in their communities and, if meeting the inclusion criterion (i.e., industrial tomato cultivation), were briefed about the survey topic and the anticipated duration of the interview, which could range between two and three hours depending on the variety of crops cultivated in each farm and the willingness of farmers to provide information. To eliminate any potential bias, it was made clear to the respondents that the study was conducted for research purposes. Participation in the survey was entirely voluntary and personal details were kept anonymous.

The questionnaire consisted of two sections. The first section aimed at recording detailed technical of economic data of the cultivation year 2020–2021 including: land use (acreage of irrigated and non-irrigated land, ownership and values/rents); annual labor requirements (working hours per activity of family members and hired personnel) as well as implicit (wages of family members that are not actually paid) and paid wages; fixed capital endowments (acquisition year, types and replacement values of buildings, mechanical equipment and land reclamation); types, quantities and variable capital used for all purchased inputs pertaining to crop production (e.g., seeds/plants, fertilizers, pesticides, fuel, irrigation water, hired machinery); variable capital used for other services (e.g., certifications, soil analysis, transportation, advisory); product yields and prices; and income support (e.g., basic payment scheme, coupled payments, agri-environmental payments) and other compensations paid to farmers. Although the collected technical and economic data refer to a specific year, farmers were requested to provide information according to their typical or usual practices, management and performance, rather than to situations associated with that particular period. Extraordinary events, such as unusual market fluctuations or natural disasters, were not taken into account in terms of the data collection process.

In the second section of the questionnaire, the extent to which the sampled farms applied SWMPs was measured. This part of the survey was based on a carefully designed list of the most typical SWMPs for controlling broomrapes, derived from a literature review [11,12] and empirical knowledge from actual practices in the study area reflecting the level of farmers’ experience and familiarity. These encompassed cover crops; crop rotation; false seeding; deep plowing; fallowing; adjustments on sowing dates; reduced tillage or no tillage; adjustments on sowing densities; competitive hybrids and cultivars; selection of competitive crops; precision weed management; mechanical weeding; and manual interventions. The selected SWMPs were rated using a 5-point Likert scale as follows: not at all (=1), low (=2), medium (=3), high (=4) and very high (=5). Agronomists and farm advisors in the study area confirmed the practicality and applicability of the presented SWMPs prior to the survey, which were also validated by researchers as potentially effective methods against broomrape parasitism.

The measurement scale of the adoption intensity of SWMPs as well as the questionnaire items recording technical and economic data were tested in a pilot survey with 15 farmers from the study area. These farmers were then excluded from the main survey.

2.3. Methodological Background and Data Analysis

The methodological framework for the analysis of the categorical (ordinal) data in this study combined: (a) A Categorical Principal Component Analysis (CatPCA) on the categorical variables (items) describing the SWMPs presented and assessed by the participants, with the objective of reducing the number of variables and generating fewer (in number) and meaningful factors (referred to as principal components or dimensions) that explain the overarching framework. (b) A Two-Step Cluster Analysis (TSCA) to develop a typology of the sampled farms according to their implemented SWMPs using the dimensions yielded by the CatPCA.

Regarding the analysis of the collected technical and economic data of farm management, basic indicators were calculated for the sampled farms and a comparative technical and economic analysis involving the “average” farm of each cluster generated by the TSCA was conducted to identify variations in economic performance.

2.3.1. Categorical Principal Component Analysis

Farmers’ responses regarding the degree to which they implemented SWMPs on their farm were evaluated using a CatPCA. This method is widely employed in multivariate analysis [47,48,49], examining the internal validity of categorical variables (i.e., ordinal and nominal variables). In particular, CatPCA is a statistical procedure aiming to convert an original group of potentially associated variables into a set of new, non-linearly correlated variables known as principal components (dimensions). The resulting set of principal components may be equal to, or smaller than the original group of variables and explains most of the information (variance) included in the original set [50]. A correlation coefficient (factor loading) is estimated in each dimension for each variable of the original set, allowing the dimension to be identified. As a rule of thumb, factor loadings greater than 0.7 indicate a correlation between the variable and the dimension; however, lower factor loadings are acceptable in social sciences. Cronbach’s-α values should be higher than 0.6–0.7 for each factor. In addition to investigating the relationship between original data and components, this method is also useful for grouping data together, thus enabling further analysis to be performed with less variables and no missing information.

2.3.2. Two-Step Cluster Analysis

The principal components generated by the CatPCA were used in the application of TSCA, by means of which the sampled farms were categorized into groups (clusters) with common characteristics in terms of their implemented SWMPs. Thus, each cluster denoted a different farm type based on the degree to which each of the SWMPs was applied. Cluster analyses are commonly used in farm typology research [51] aiming to determine relatively homogeneous groups that may be distinguished from each other. TSCA is an extension of a typical cluster analysis enabling the identification of clusters that share common features based on categorical and/or continuous variables, which is a basic difference between the TSCA and other clustering methods [50]. The number of clusters can either be pre-defined by the researcher or calculated automatically by the algorithm. At the first step, a set of initial clusters or “pre-clusters” is created through a quick clustering algorithm scanning the data case by case and selecting whether each case should be added to a previously formed “pre-cluster” or start a new one. After the pre-clustering is complete, the “pre-clusters” are merged into larger clusters in the second step using the standard hierarchical clustering algorithm [50].

2.3.3. Comparative Technical and Economic Analysis

Using the data recorded in the first section of the questionnaire, basic technical and economic indicators were calculated for the sampled farms through the application of a technical and economic analysis. This method is the core of agricultural economics research and serves as the principal technique for the appraisal of a farming system [52] with the goal of providing a thorough description of the organization and economic performance of farms. As part of this analysis, technical indicators related to farm organization, economic indicators presenting the composition of farm revenues per income source and the structure of farm expenses per production factor (land, labor, capital) as well as basic economic results, were calculated. The latter constitute fundamental indicators of economic performance, reflecting the long-term economic viability of farms and they are outlined below:

• Gross revenue. The sum of the values of all products sold on the market plus income support payments.
• Production expenses. The sum of land use expenses (implicit rent of owned land and paid rent of rented land); labor expenses (implicit wages of family members and paid wages of hired personnel); fixed capital expenses (depreciations, interests, premiums and maintenance of buildings, mechanical equipment and land reclamation); and variable capital expenses (including the value of all purchased inputs and other services).
• Net profit/loss. The difference between gross revenue and production expenses.
• Return to labor. The algebraic sum of labor wages (implicit and paid) plus net profit/loss divided by total working hours.
• Farm income. The algebraic sum of land rent (paid and implicit); labor wages (paid and implicit); capital interest; and net profit/loss.
• Gross margin. The difference between gross revenue and variable capital expenses.

Based on the farm typology developed by combining the CatPCA with the TSCA, the main indicators and economic results of an “average” farm of each farm type (i.e., the farm representing the mean of all farms classified within each type) were then estimated and compared (comparative technical and economic analysis), revealing discrepancies in structural organization, production costs and revenues as well as in overall economic performance among the different farm types. Thus, the analysis indicated the economic potential of farms in relation to the use of SWMPs.

2.3.4. Statistical Analysis

The SPSS 24.0 statistical package was used for the statistical analysis, which involved both parametric and non-parametric methods. In particular, one-way ANOVA was carried out to compare the means of the dimensions extracted by the CatPCA and the Bonferroni post hoc test was applied to evaluate the normality and homogeneity of the data. However, in cases where the elements of the data did not follow a normal distribution, the results of one-way ANOVA were not accepted and the non-parametric Mann–Whitney U test was used to evaluate differences between specific groups.

3. Results

3.1. Farm Typology

The 13 initial variables describing the SWMPs presented and evaluated by the sampled farmers were categorized by the CatPCA into four dimensions with eigenvalues greater than 1, explaining 80.09% of the total variance (

Table 1. Rotated Categorical Principal Component Analysis (CatPCA) table: loadings of variables describing Sustainable Weed Management Practices (SWMPs).

SWMPs	Dimensions
	Component Loadings *
	1	2	3	4
Cover crops	−0.139	−0.560	0.682	−0.162
Crop rotation	0.712	0.256	−0.056	−0.112
False seeding	0.837	−0.313	−0.196	0.148
Deep plowing	0.340	0.643	0.234	−0.312
Fallowing	0.187	0.557	0.548	0.508
Adjustments on sowing dates	0.867	−0.276	−0.114	−0.012
Reduced tillage or no tillage	0.826	−0.353	−0.262	0.113
Adjustments on sowing densities	0.859	−0.201	−0.193	−0.059
Competitive hybrids and cultivars	0.025	−0.548	0.702	−0.271
Selection of competitive crops	0.478	−0.346	0.676	−0.113
Precision weed management	0.337	0.320	0.462	0.694
Mechanical weeding	0.364	0.674	0.106	−0.368
Manual interventions	0.336	0.622	0.185	−0.378
Cronbach’s-α (rotated matrix)	0.822	0.697	0.588	0.236
Eigenvalue	4.138	2.807	2.188	1.280
% of variance explained	31.83	21.59	16.83	9.84
% of variance explained (Total)	80.09

* Component loadings in bold are the highest and most important in each dimension.

). Dimension 1 was characterized as “Cultural”, due to the fact that the five variables with the highest loadings focused on cultivation techniques (crop rotation, false seeding, adjustments on sowing dates, reduced tillage or no tillage and adjustments on sowing densities). In dimension 2, the highest loadings were detected for variables that referred to deep plowing, mechanical weeding and manual interventions. Thus, it was titled “Physical–mechanical”. Dimension 3 was named “Crop–cultivar selection”, as the four variables with the highest loadings described SWMPs such as cover crops, fallowing, competitive hybrids and cultivars as well as selection of competitive crops. In dimension 4, one variable exhibited the highest loading associated with precision weed management methods. Consequently, it was designated as “Precision agriculture”. The low Cronbach’s-α value in this case was mainly due to the fact that this dimension was characterized by a single variable of the original set and thus does not necessarily indicate poor reliability. “Precision agriculture” was retained in the analysis, as it reflects a contemporary approach that is conceptually significant for distinguishing farms based on the use of SWMPs. Therefore, its impact on the farm typology and the ensuing technical and economic analysis remained theoretically substantiated yet interpreted with caution.

The TSCA yielded four clusters, each representing a different farm type, in which the 76 sampled farms were grouped (

Table 2. Results of the Two-Step Cluster Analysis (TSCA): alternative farm types.

Clusters	Cluster Size		Dimensions
			“Cultural”		“Physical– Mechanical”		“Crop–Cultivar Selection”		“Precision Agriculture”
	Number of Farms	Percentage (%)	Mean	St. Dev.	Mean	St. Dev.	Mean	St. Dev.	Mean	St. Dev.
1	33	43.4	0.391 a	1.221	−0.367 a	0.430	−0.736 a	0.451	0.319 a	0.741
2	2	2.6	2.049 b	1.128	1.949 b	2.307	2.808 b	1.162	4.224 b	1.191
3	21	27.6	−0.267 c	0.286	1.245 b	0.289	−0.179 c	0.402	−0.619 c	0.414
4	20	26.3	−0.570 c	0.457	−0.896 c	0.357	1.122 d	0.535	−0.299 c	0.497
Total	76	100.0	0.000	1.006	0.000	1.006	0.000	1.006	0.000	1.006

Note: Means in the same column denoted with same superscripts are not statistically significantly different (Bonferroni post hoc test) at the 10% significance level.

). The first cluster, which was the largest, included 33 farms (43.4% of the total sample), the second 2 (2.6%), the third 21 (27.6%) and the fourth 20 (26.3%). Clustering was performed based on the four dimensions derived from the CatPCA, as their effect was significant in the formation of the four clusters. The specific elements characterizing each cluster can be ascertained by investigating the means of each dimension and their corresponding signs, together with the frequency analysis of each individual variable comprising the four dimensions describing the selected SWMPs (

Table 3. Frequency analysis (%) of variables describing Sustainable Weed Management Practices (SWMPs).

Variables	Clusters				Total Sample
	Cluster 1 “Transitional”	Cluster 2 “Holistic”	Cluster 3 “Conventional”	Cluster 4 “Selective”
Dimension 1 “Cultural”
- Crop rotation
Not at all	6.1	0.0	0.0	0.0	2.6
Low	12.1	0.0	9.5	15.0	11.8
Medium	24.2	0.0	14.3	50.0	27.7
High	12.1	0.0	47.6	30.0	26.3
Very high	45.5	100.0	28.6	5.0	31.6
- False seeding
Not at all	33.3	0.0	85.7	70.0	56.6
Low	12.1	50.0	14.3	25.0	17.1
Medium	15.2	0.0	0.0	5.0	7.9
High	24.2	0.0	0.0	0.0	10.5
Very high	15.2	50.0	0.0	0.0	7.9
- Adjustments on sowing dates
Not at all	6.1	0.0	57.2	35.0	27.6
Low	15.1	0.0	19.0	30.0	19.7
Medium	18.2	50.0	14.3	10.0	15.8
High	21.2	0.0	9.5	20.0	17.1
Very high	39.4	50.0	0.0	5.0	19.8
- Reduced tillage or no tillage
Not at all	24.2	50.0	81.0	50.0	47.4
Low	9.1	0.0	14.3	25.0	14.5
Medium	12.1	0.0	4.7	25.0	13.1
High	36.4	0.0	0.0	0.0	15.8
Very high	18.2	50.0	0.0	0.0	9.2
- Adjustments on sowing densities
Not at all	6.0	0.0	9.5	0.0	5.3
Low	18.2	0.0	19.0	30.0	21.0
Medium	9.1	50.0	38.1	55.0	30.3
High	27.3	0.0	28.6	15.0	23.7
Very high	39.4	50.0	4.8	0.0	19.7
Dimension 2 “Physical– mechanical”
- Deep plowing
Not at all	15.1	0.0	0.0	5.0	7.9
Low	45.4	0.0	0.0	20.0	25.0
Medium	21.2	0.0	14.3	60.0	28.9
High	9.1	50.0	57.1	10.0	23.7
Very high	9.1	50.0	28.6	5.0	14.5
- Mechanical weeding
Not at all	0.0	0.0	0.0	0.0	0.0
Low	24.2	0.0	0.0	25.0	17.1
Medium	39.4	0.0	0.0	50.0	30.3
High	21.2	50.0	28.6	20.0	23.7
Very high	15.2	50.0	71.4	5.0	28.9
- Manual interventions
Not at all	0.0	0.0	0.0	0.0	0.0
Low	18.2	0.0	0.0	25.0	14.5
Medium	42.4	0.0	14.3	45.0	34.2
High	24.2	50.0	28.6	25.0	26.3
Very high	15.2	50.0	57.1	5.0	25.0
Dimension 3 “Crop-cultivar selection”
- Cover crops
Not at all	60.6	50.0	85.7	0.0	51.3
Low	24.3	0.0	9.5	5.0	14.5
Medium	12.1	50.0	4.8	15.0	11.8
High	0.0	0.0	0.0	60.0	15.8
Very high	3.0	0.0	0.0	20.0	6.6
- Fallowing
Not at all	69.7	0.0	4.8	60.0	47.4
Low	24.2	0.0	28.6	20.0	23.7
Medium	6.1	0.0	57.1	15.0	22.4
High	0.0	50.0	9.5	5.0	5.2
Very high	0.0	50.0	0.0	0.0	1.3
- Competitive hybrids and cultivars
Not at all	48.5	0.0	42.9	0.0	32.9
Low	21.2	50.0	19.0	0.0	15.8
Medium	12.1	0.0	28.6	0.0	13.2
High	12.1	50.0	9.5	40.0	19.7
Very high	6.1	0.0	0.0	60.0	18.4
- Selection of competitive crops
Not at all	18.2	0.0	23.8	0.0	14.5
Low	15.1	0.0	28.6	0.0	14.5
Medium	27.3	0.0	33.3	10.0	23.7
High	21.2	0.0	14.3	45.0	25.0
Very high	18.2	100.0	0.0	45.0	22.3
Dimension 4 “Precision agriculture”
- Precision weed management
Not at all	78.8	0.0	100.0	75.0	81.6
Low	18.2	0.0	0.0	5.0	9.2
Medium	0.0	0.0	0.0	20.0	5.3
High	3.0	0.0	0.0	0.0	1.3
Very high	0.0	100.0	0.0	0.0	2.6

Note: The four dimensions generated by the Categorical Principal Component Analysis (CatPCA) are indicated in bold.

). In what follows, a concise overview of the four clusters is provided.

The first cluster was characterized by the low positive sign of the means of the dimensions “Cultural” and “Precision agriculture” and the negative sign of the means of “Physical–mechanical” and “Crop–cultivar selection”, implying that the farms included in this cluster did not strongly align with any particular dimension. In fact, these farms exhibited a modest tendency to use cultural practices such as crop rotation (45.5% very high; 12.1% high), adjustments on sowing dates (39.4% very high; 21.2% high), reduced tillage or no-tillage (18.2% very high; 36.4% high) and adjustments on sowing densities (39.4% very high; 27.3% high) but lacked commitment to a defined approach. This type depicted farms in a stage of strategic uncertainty, with weed management decisions remaining adaptable, fluid or under investigation. Therefore, they were labeled “Transitional”.

The second cluster was distinctly identified by the high positive sign of the means across all four dimensions. Thus, this cluster comprised farms that employed a variety of SWMPs as part of an overall, quite diverse approach. Notably, 100.0% of this cluster reported very high use of SWMPs including crop rotation, selection of competitive crops and precision weed management, with the latter especially receiving by far the highest score among the four clusters. In addition, SWMPs such as deep plowing, mechanical weeding, manual interventions and fallowing were also applied at levels that exceeded those observed in other clusters (50.0% very high; 50.0% high) and adjustments on sowing dates and densities to an extent that was considerable (50.0% very high; 50.0% medium). As can be seen, the farms classified in this cluster adopted a broad array of SWMPs, grounded in both traditional and contemporary agronomic knowledge and they were thus called “Holistic”.

The third cluster was distinguished by the negative sign of the means of the dimensions “Cultural”, “Crop–cultivar selection” and “Precision agriculture”, whereas “Physical–mechanical” exhibited a positive mean, describing the farms grouped in this cluster. Their common characteristics involved the extensive use of crop rotation (28.6% very high; 47.6% high), deep plowing (28.6% very high; 57.1% high), mechanical weeding (71.4% very high; 28.6% high) and manual interventions (57.1% very high; 28.6% high). On the contrary, they hardly utilized SWMPs such as false seeding (85.7% not at all; 14.3% low), adjustments on sowing dates (57.2% not at all; 19.0% low), reduced tillage or no-tillage (81.0% not at all; 14.3% low) and cover crops (85.7% not at all; 9.5% low), while none of the farms of this cluster reported using precision weed management (100.0% not at all). This type reflected farms primarily relying on well-established SWMPs, rooted in traditional, long-standing farming principles rather than on novel approaches. Due to this fact, they were assigned the title “Conventional”.

The fourth cluster was delineated by the positive sign of the mean of the dimension “Crop–cultivar selection”, which defined the farms of this cluster, while the dimensions “Cultural”, “Physical–mechanical” and “Precision agriculture” demonstrated negative means. These farms were particularly engaged in SWMPs such as selection of competitive crops (45.0% very high; 45.0% high), cover crops (20.0% very high; 60.0% high) as well as competitive hybrids and cultivars (60.0% very high; 40.0% high), with the latter two scoring the highest among the four clusters. A neutral attitude was expressed regarding the use of crop rotation (30.0% high; 50.0% medium; 15.0% low), mechanical weeding (20.0% high; 50.0% medium; 25.0% low) and manual interventions (25.0% high; 45.0% medium; 25.0% low), while SWMPs that were not found popular included false seeding (70.0% not at all; 25.0% low), reduced tillage or no-tillage (50.0% not at all; 25.0% low) and precision weed management (75.0% not at all; 5.0% low). As their approach was essentially delicate, emphasizing prevention through careful crop and cultivar selection combined with the use of cover crops and competitive hybrids to naturally suppress weeds while avoiding active intervention, they were given the label “Selective”.

3.2. Farm-Level Economic Effects of the Four Approaches

The calculation of technical indicators revealed important differences across farm types regarding structural and organizational aspects (

Table 4. Technical and economic indicators of farm types according to the implemented Sustainable Weed Management Practices (SWMPs) as well as of the average farm of the total sample.

Indicators	Farm Types				Total Sample (n = 76)
	“Transitional” (n = 33)	“Holistic” (n = 2)	“Conventional” (n = 21)	“Selective” (n = 20)
Farm area (ha)	70.5	86.5	62.9	50.3	63.5
Cotton	9.6	2.5	15.7	7.4	10.5
Winter cereals	36.4	37.5	27.1	24.9	30.9
Industrial tomato	12.6	26.5	9.6	8.0	10.9
Others (legumes, maize, oilseed rape)	11.9	20.0	10.5	10.0	11.2
Labor requirements (hours/ha/year) 1	61.3	68.0	79.2	59.8	66.1
Family	42.3	43.0	54.0	46.0	46.3
Hired	19.0	25.0	25.2	13.8	19.8
Gross revenue (€/ha)	3355.3	4241.5	3621.5	3200.4	3427.7
Product sales (weighted average)	2601.3	3498.9	2842.2	2499.7	2678.3
Industrial tomato	7731.0	7900.4	7808.4	7503.7	7716.7
Income support	754.0	742.6	779.3	700.7	749.4
Production expenses (€/ha)	3312.8	3753.0	3601.5	3284.3	3401.7
Land expenses	499.5	564.3	513.6	488.4	503.4
Labor expenses	214.7	244.2	283.6	212.9	234.2
Fixed expenses	892.3	958.5	1136.8	932.7	970.1
Variable expenses	1706.3	1986.0	1667.5	1650.3	1694.0

¹ Including working hours spent on general tasks (e.g., information/communication, equipment maintenance, marketing, management supervision). Note: The basic indicators are presented in bold. The revenues from industrial tomato products are indicated in italics.

). “Holistic” farms were the largest in size, occupying an average farm area of 86.5 ha, while “Selective” farms represented the smallest type, with an average farm area of 50.3 ha. Winter cereals were cultivated to a greater extent than all other crops across the four farm types, surpassing the half of total farm area in “Transitional” farms (51.6%). “Holistic” farms allocated a substantial percentage of their farm area to other crops including collectively maize, legumes and oilseed rape (23.1%), while cotton stood for a significant share of total farm area in “Conventional” farms (24.9%). The proportion of total farm area devoted to industrial tomato ranged significantly from 15.3% in “Conventional” to 30.6% in “Holistic” farms. With regard to labor, “Conventional” farms exhibited the highest annual requirements (79.1 h/ha), whereas “Selective” farms the lowest (59.8 h/ha). An important difference was also observed in the synthesis of labor sources, as 31.0% of total labor requirements were covered by hired personnel in “Transitional” and “Conventional” farms, with this percentage rising to 36.0% and declining to 23.0% in “Holistic” and “Selective” farms, respectively. Information regarding the labor requirements of each crop is included in Appendix A (

Table A1. Labor requirements per crop of farm types according to the implemented Sustainable Weed Management Practices (SWMPs) as well as of the average farm of the total sample.

Labor Requirements (hours/ha/year) 1	Farm Types				Total Sample (n = 76)
	“Transitional” (n = 33)	“Holistic” (n = 2)	“Conventional” (n = 21)	“Selective” (n = 20)
Cotton	88.5	113.2	100.4	88.9	92.8
Winter cereals	30.7	37.2	37.9	25.1	31.7
Industrial tomato	84.1	85.1	100.4	80.3	87.3
Legumes (for human consumption)	41.1	59.7	49.6	30.8	41.6
Legumes (for forage)	30.8	34.2	36.6	26.1	31.6
Maize	80.0	96.6	91.9	76.2	83.3
Oilseed rape	39.1	44.4	44.7	35.0	39.9
Total (weighted average)	51.4	59.3	66.8	46.2	54.8

¹ Not including working hours spent on general tasks (e.g., information/communication, equipment maintenance, marketing, management supervision).

).

Economic indicators per ha are also reported in Table 4. “Holistic” farms achieved the highest gross revenue among the four types, as a result of obtaining the highest weighted average income from all product sales, along with the highest income from industrial tomato in particular. Due to this fact, these farms ranked as the most productive, followed by “Conventional” farms, which received the highest amount of income support payments. The contribution of income support to gross revenue was roughly equal across “Transitional”, “Conventional” and “Selective” farms, varying between 21.5% and 22.5% of total revenues, while “Holistic” farms demonstrated a lower rate of 17.5%. Despite attaining the highest gross revenue and being the most market-oriented, “Holistic” farms incurred the highest production expenses, which was mainly attributed to their high variable costs. On the contrary, “Selective” farms had the lowest production expenses, as they operated with the lowest costs associated with land use, labor and variable capital. Although varying greatly (from 1650.3 €/ha in “Selective” to 1986.0 €/ha in “Holistic” farms), variable expenses accounted for the highest share of total costs in all farm types. Fixed expenses were rather low in most types, except for “Conventional” farms, which were particularly oriented to mechanical weeding and had invested in relevant machinery. As expected, these farms also reported the highest labor expenses due to their high labor requirements.

Significant differences among the four approaches were detected in terms of their economic results, thus reflecting the economic dynamics of SWMPs (Figure 2). Despite significant variation (from 706.8 €/ha in “Selective” to 1383.4 €/ha in “Holistic” farms) which implies improvement potential, the farm income was positive in all farm types, highlighting the capacity of industrial tomato sector to provide adequate income and guarantee the livelihoods of rural families in Greece. The positive sign also shows that the utilization of production factors (land, labor and capital) was remunerated sufficiently. Farm income is generally regarded as an economic result that offers an integrated depiction of economic performance. The gross margin was also positive in all farm types, revealing a satisfactory financial status in the short run, with sufficient liquidity to address labor and fixed expenses. Nevertheless, the net profit was very low in “Transitional” (42.5 €/ha) and “Conventional” (20.0 €/ha) farms, while “Selective” farms operated with a net loss (−83.9 €/ha). This implies that these three types lack economic viability in the long run and thus require radical operational modifications. “Holistic” farms constituted the only exception, as they achieved remarkable net profits (488.5 €/ha) due to their high productivity. These farms exhibited also an impressive return to labor (10.8 €/hour), which was somewhat acceptable in “Transitional” (4.2 €/hour) and “Conventional” farms (3.8 €/hour). On the contrary, “Selective” farms demonstrated a quite low return (2.2 €/hour), actually lower than the minimum hourly wage in Greece.

4. Discussion

Despite some commonalities, the comparative technical and economic analysis revealed important differences across the four farm types in terms of organizational characteristics and economic performance. These differences prompt intriguing questions regarding the economic impact of SWMPs at the farm level and their potential to counterbalance possible income losses with cost savings and thus ensure farm profitability. Are SWMPs associated with higher economic performance when implemented to a greater extent? Is there an “optimal” combination of SWMPs which guarantees economic viability in the long run? If so, what are the prerequisites for successful business models and how do they translate into policy implications? This Section discusses the relationship between the SWMPs for broomrape control and the economic performance of industrial tomato farms in the Region of Thessaly in Central Greece while suggesting targeted interventions based on farm typology.

This study confirmed the importance of considering farm heterogeneity and employing a typology-based approach for classifying and characterizing farms according to their implemented SWMPs, as farm types differed not only in structure but also in revenues, expenses and profitability. This is in line with previous typological studies which pinpointed disparities in many aspects related to farm management including land and input use, labor sources and market orientation across different farm types implementing sustainable practices [53,54,55]. Through different levels of labor and capital utilization, each of the four identified types exhibited distinct organizational requirements and involved variations in management and daily operation. When analyzed from the perspective of entrepreneurship, however, they were found to operate either with net losses or nearly negligible net profits. Indeed, the low economic performance of arable farms in Greece is a prevalent observation in the recent literature, raising concerns about their long-term economic viability [56,57]. However, “Holistic” farms deviated from this pattern, in accordance with recent reviews arguing that holistic approaches are particularly beneficial in terms of enhancing productivity, as they contribute to alleviating the risk and uncertainties associated with the efficacy of a single weed control strategy [8,22,58,59,60]. Prior work focusing on the evaluation of technical performance has also shown that the integration of multiple methods reduces anticipated and actual yield losses due to weed occurrence while diminishing reliance on chemical tools [61,62,63]. Nevertheless, the farm income and, to a greater extent, the gross margin of all farm types in this study were quite satisfactory, corroborating similar results of Finco et al. [64] and Kouriati et al. [65], who also reported positive values. This showcases that Mediterranean farms engaged in arable production possess future potential contingent upon certain conditions. These mainly pertain to cost reduction, which relies on more efficient resource utilization, or to product valorization through higher selling prices.

Our findings indicated a strong correlation between the adoption degree of SWMPs and economic outcomes. Contrary to expectations, however, no evidence suggested that higher adoption levels are linked to lower economic performance. “Holistic” farms, which exhibited the highest levels of implementation, incurred increased costs, especially in terms of inputs, yet attained higher productivity, resulting in improved profitability and economic resilience. “Transitional” and “Selective” farms, on the other hand, were characterized by lower adoption rates and achieved lower production expenses, but remained more susceptible to broomrape infestation, implying lower productivity and, thus, uncertain profits and returns. “Conventional” farms also employed a limited range of SWMPs, as they were solely committed to mechanical and physical methods. In this farm type, nonetheless, the low economic performance was mostly attributed to elevated fixed capital and labor costs rather than to low productivity. While mechanical and physical weed control may have significant crop yield potential, they can be subject to limited profitability due to requirements for additional machinery and labor [66,67]. Evidently, these farms are in need of substantial internal restructuring, as they appear to have undertaken irrational investments burdening their expenses. In spite of their high revenues, their excessive production costs limit their performance and competitiveness and render them more dependent on income support.

In light of the multitude of SWMPs available for broomrape control, farmers often experience confusion in selecting the most effective and applicable strategies for their farms as well as their combinations, especially given the exceptional challenges associated with the management of parasitic weeds [68]. Our study identified a holistic approach that can have a neutral or even positive impact on farm economic performance. This entails a strategic combination of preventive SWMPs including crop rotation, modifications of sowing dates and densities, mechanical cultivation or tillage, selection of competitive, resistant or transgenic crops and curative strategies such as hand weeding. Previous research has noted that combining such measures effectively within an integrated broomrape management framework, with a particular emphasis on preventive tactics, can potentially yield significant agronomic and environmental benefits [4,12,69]. Therefore, our study substantiates the sustainability of such interventions from an economic standpoint. Moreover, we demonstrated the key role of precision weed management as an element of this “optimal” combination of SWMPs. Particularly, “Holistic” farms acquired unmanned aerial vehicles (UAVs), commonly known as drones, for remote sensing aiming at enhancing detection of broomrape infestation throughout their entire area. Although the adoption cost of UAV-based monitoring methods may vary depending on their technical characteristics, it is still much lower than that of similar precision technologies [70,71,72]. This is indeed reflected in the fixed costs of “Holistic” farms, which remained at moderate levels. Our outcomes can therefore contribute to the promotion of precision weed management, which was found to be underutilized in the study area. The current scarcity of farms implementing a highly diverse and holistic approach involving precision tools was indicated by this small cluster, the size of which is expected to increase in the upcoming years. The minimal adoption rates of precision weed management recorded in “Transitional” and “Selective” farms pertain to the use of digital applications for monitoring agrochemical levels.

Consequently, the achievement of this “optimal” combination of SWMPs was not uniform across farm types, underscoring the need for favorable external conditions relating to labor availability, income support policies, market dynamics and advisory support. In particular, “Holistic” farms coped with increased production expenses and capitalized on the benefits of SWMPs, as they were more market-oriented and had adequate resources and better access to knowledge. The latter was evident not only in their effective application of SWMPs leading to high productivity but also in their proper resource utilization. Indeed, their high production costs mostly stemmed from their particular engagement in tomato cultivation, an input-intensive crop [73], rather than from labor mismanagement and/or irrational investments in fixed capital during the adoption process of SWMPs, which was seen in other farm types. Therefore, the role of farm advisory is crucial, especially in terms of assisting farmers during their transition to such practices. Considering the EU’s supporting policy in the context of Agricultural Knowledge and Innovation Systems (AKIS), which brings together advisors, researchers and farmers [74], targeted advisory involving the establishment of demonstration farms is required toward guiding farmers on managerial decision-making throughout the adoption process, providing them with the requisite skills and enhancing their confidence when implementing SWMPs.

With regard to labor, the adoption of SWMPs characterized by high labor intensity is quite challenging at this time, especially under the pressure of negative rural population dynamics and declining availability of hired personnel prevailing in Greek agriculture [75]. According to the findings of our study, at least 30.0% of labor was hired almost in all farm types, except for “Selective” farms, which accounted for the least labor-intensive type. This actually indicates the dependency of farms on external workers, whose presence is irregular, but also whose employment burdens production expenses, as demonstrated by the examples of “Holistic” and “Conventional” farms. Structural policies including generational renewal initiatives are needed to increase the involvement of family members, which could eventually provide more profitable directions for Greek and Mediterranean arable farming. Such measures are reflected in the EU’s vision for a thriving agricultural sector, which includes the attractiveness of rural areas and farming sector for future generations as a means of enticing them to remain in the profession [76].

Income support exhibited no considerable variation across the farm types and did not constitute a major part of revenues. Indeed, all the four types in this study were market-integrated, as over 77.5% of their total revenues came from markets. This percentage in a recent study in Greece was not found to be higher than 66.0% for arable farms specializing in cereal and legume production [77], highlighting the significant market potential of industrial tomato. On the other hand, the technical and economic analysis here has shown that income support was crucial for the long-term survival of all farm types, as its abolition would result in severe net losses for all, including “Holistic” farms. Targeted and fair measures are thus necessary for addressing additional costs and offsetting potential income losses, especially during the transition period. This need mostly pertains to “Transitional”, “Conventional” and “Selective” farms. Ziehmann et al. [78] found evidence that a cost-effective strategy for promoting non-chemical alternatives involves the provision of an equal amount of per-hectare payments to all farmers to mitigate transition-related risks by ensuring a stable income source, alongside price markups for higher yields per hectare to encourage the selection of practices that maintain production levels.

Reliance on policy-related payments may indeed compromise farm economic performance and, in turn, long-term economic viability, unless paired with market integration [79]. In this regard, markets can provide farmers with commercial opportunities by incentivizing them to adopt SWMPs under certain conditions. Understanding consumer preferences toward SWMPs and identifying consumer clusters with demand for sustainably produced food can offer valuable insights to farmers who may be unaware of possible market outlets for their products or perceive SWMPs merely as a “necessary evil” to fulfill regulatory or societal expectations. Certification, labeling and the establishment of social networks with knowledge exchange platforms analyzing consumer trends in the food market can be pivotal for farmers toward this direction.

5. Conclusions

SWMPs are constantly gaining attention due to their potential agronomic and environmental benefits. However, challenges linked to low productivity and increased production expenses remain barriers to on-farm adoption. Motivated by the need for effective parasitic weed control in Mediterranean cropping systems, this study employed a typology-based approach to determine the impact of SWMPs against broomrape parasitism on the structural organization, management and economic performance of industrial tomato farms in a typical agricultural district of Central Greece. A combination of PCA with TSCA yielded four distinct farm types according to their implemented SWMPs, namely “Transitional”, “Holistic”, “Conventional” and “Selective”. The findings of a comparative technical and economic analysis indicated that varying implementation levels of SWMPs were reflected in resource utilization and managerial decisions, thereby leading to considerable disparities in economic performance across farm types. “Holistic” farms, which exhibited the highest adoption levels, implemented an effective broomrape control strategy and achieved a remarkable net profit of 488.5 €/ha, unveiling the economic potential of SWMPs under certain external conditions. The outcomes of this study enhance the practical relevance of the farm typology developed, offering useful insights for farm-level decision making and tailored advisory based on the particular needs and characteristics of each farm type. Farmers and advisors could benefit from these findings to make more informed decisions regarding the use of SWMPs and identify more “optimal” combinations of practices that ensure economic viability.

To the best of our knowledge, this study is the first to estimate the effects of different combinations of SWMPs on farm economic performance, and it validates earlier observations of research emphasizing technical aspects. Nevertheless, it provides a comprehensive overview of the nuanced trade-offs between the costs and benefits associated with the on-farm adoption of SWMPs in a real-world Mediterranean context, confirming that the economic effects of such practices can be better understood through an analysis at the farm level, where adoption-related managerial decisions are explicitly reflected. Considering the lack of a detailed database of technical and economic data, such as the Farm Accounting Data Network (FADN) of the EU, for farms already implementing SWMPs or in the transition process, its findings also yield practical implications at the policy level. Consultants and policymakers could utilize them as a starting point for assisting farms that face specific challenges as well as for promoting new or established successful business models. Creating a common database that also includes information for intertemporal variations in yields and input use could facilitate the design of targeted policies and measures for a broader adoption, further development and evolution of SWMPs. This study proposes specialized advisory services involving the establishment of demonstration farms to assist farmers during the learning process of SWMPs, generational renewal initiatives to reinforce the presence of young farmers and secure labor availability in rural areas and, last but not least, targeted income support payments to address transition-related risks combined with specific market integration incentives. The need to integrate SWMPs that minimize external inputs in particular is evident in the trends of input costs and product prices. Since the cultivation year 2020–2021, to which the data of this study refer, the prices of variable inputs increased on average by 22.9%, driven by increases in the costs of energy, agrochemicals and fertilizers. During the same period, product prices of cereal and industrial crops decreased on average by 7.5% and 19.6%, respectively [80]. Therefore, less reliance on such inputs can substantially improve farm resilience.

Future research based on more integrated approaches incorporating social, economic and environmental dimensions in modelling exercises could ensure more holistic substantiations regarding the adoption of SWMPs in arable farms, while considering spatial variations in broomrape infestation and diverse agroecological and farming conditions at the local/regional level.