Decision Support System for the Assessment and Enhancement of Agrobiodiversity Performance

Abstract

The industrialisation of agriculture and changes in production systems have contributed to a biodiversity decline worldwide. Developing accurate and adequate assessment methods can encourage farmers to support more sustainable agricultural management. This study presents a decision support system to promote agrobiodiversity that incorporates not only a quantitative assessment of relevant indicators of agrobiodiversity performance but also provides enhancement practice recommendations and associated benefits, presenting an action plan in order of priority. Additionally, the decision support system allows a visual comparison between biodiversity composite indicators and indicators representing pest control and crop yield. Since grape cultivation is considered one of the most intensive agricultural systems, thus significantly impacting biodiversity, the elaborated decision support system was tested on a viticultural agroecosystem in the demarcated Douro region in Portugal. The results demonstrated the decision support system functioning according to the selected methodology and allowed the identification of future lines for investigation. During the analysed period, the following were verified: an increase of 2% in the biodiversity indicator, 130% in harvest yield, and 2077% in the enemy-to-pest ratio. It is expected that the elaborated DSS will offer a significant contribution by bridging communication gaps on alternative management options to improve biodiversity performance in agricultural systems.

1. Introduction

Agrobiodiversity covers the diversity of living organisms (plants, animals, microorganisms, among others) with relevance to agriculture [1]. Functional agrobiodiversity, in particular, corresponds to the beneficial elements of biodiversity for the provision of agricultural goods [2,3]. The wide range of elements composing agrobiodiversity can be clustered into two main categories: planned biodiversity and unplanned biodiversity [4]. Planned biodiversity encompasses the spatial and temporal arrangement of elements that are intentionally introduced and maintained in a certain agroecosystem (such as crops and livestock), whereas unplanned biodiversity is composed of all soil biota, decomposers, carnivores, herbivores, and all the living organisms present in the agroecosystem [4,5]. Both types of diversity can significantly impact productivity, soil health, pest regulation, and other important ecosystem services [4], conditioning the sustainability of agricultural systems [1]. However, the industrialisation of agriculture and changes in production systems, namely the tendency to monoculture or few crop varieties, some soil management techniques, and the large-scale use of synthetic pesticides and fertilisers, have been relevant causes of biodiversity decline worldwide [6]. The Food and Agriculture Organization (FAO) of the United Nations revealed that, in 2019, 36.47% of the terrestrial surface was allocated to agriculture [7]. The large proportion of terrestrial surface occupied by modern, intensive, and conventional agriculture contributes to reducing ecological infrastructures to maintain several organisms [8]. Furthermore, using chemical pesticides leads to escalated levels of soil, air, and water pollution, which, in turn, damages non-target species that benefit agroecosystems [9,10]. Inorganic fertiliser use is also associated with changes in regular ecosystem functioning affecting the diversity of microorganisms, mammals, and birds, among other organisms [11]. The population growth and the high per capita consumption of natural resources in the industrialised world are also responsible for the rapid loss of agrobiodiversity [5].

There is ample evidence that despite ongoing efforts, biodiversity is declining at unprecedented rates in human history [12]. The 2023 FAO report on the progress towards Sustainable Development Goals (SDGs) highlights the deterioration of global trend assessment in three out of five indicators associated with the 15th SDG “Life on land”, namely forest area as a proportion of total land (SGD 15.1.1.), proportion of land that is degraded over the total land area (SDG 15.3.1.), and proportion of degraded mountain land (15.4.2 (b)) [13]. Such results support the need for reflection on what should be changed to improve the efficacy of corrective actions to stop, or at least to reduce, biodiversity loss, by farmers, who are the sole decision makers and managers in farming operations [14].

Given the impact of some agriculture techniques and management on biodiversity, agricultural companies must develop strategies to improve their sustainable performance [15]. The term sustainability etymologically derives from the Latin verb “sustinere”, which means maintaining and ensuring that a certain thing lasts, as well as committing in this sense [16]. Thus, sustainability corresponds to the preservation of something that exists in the present and that must be maintained in the future, involving accountability and commitment [16]. In this study, the term “sustainable” is used to describe agricultural practices and/or values of indicators that are aligned with recent recommendations (by scientific literature and reputable organisations) to ensure the conservation of functional biodiversity in agricultural management. Assessment methods can support farmers in their management activities by providing detailed information including, enhancement practices, weaknesses, and strengths [17]. According to Siebrecht [17], more investigation is required into how theory can be applied to promote sustainability in agriculture. When it comes to evaluating the performance of agrobiodiversity, previous studies have mainly focused on topics such as genetic diversity [18,19,20,21,22], conservation status or population assessments [23,24], dietary diversity [25,26,27,28,29], and public policies [30]. There is a need to create methodologies aiming to assess agrobiodiversity comprehensively and that are more practical and able to bridge useful information gaps [31,32]. A survey performed with companies with agricultural supply chains revealed that biodiversity impact assessment is associated with concerns of how to measure and how to aggregate results, among others. Another survey found that investments in biodiversity are limited by the difficulty of measuring it, a lack of knowledge, and a lack of available data [33]. A Decision Support System (DSS) providing an action plan can bridge communication gaps and enable decision makers in the agricultural sector to adopt innovative practices to foster biodiversity performance [32].

Some currently common methodologies to assess biodiversity in agricultural systems are characterised by difficult interpretation [34]. It is recommended to provide more transparent assessment methods that allow a clear understanding of what actions should be implemented, encouraging decision makers to support biodiversity-friendly agriculture [35,36]. A DSS providing clear recommendations can address the limitations of current methodologies such as the difficulty in interpreting indicator values and identifying priority action areas [31,32].

This study aims to present a DSS for agrobiodiversity assessment that incorporates a quantitative assessment of relevant indicators of biodiversity performance and provides enhancement practice recommendations. It is also intended to test the DSS’s functioning and compare the evolution of overall biodiversity performance indicators with pest control and crop yield indicators. It is expected that the elaborated DSS will offer a significant contribution by bridging communication gaps on alternative management options to improve biodiversity performance in agricultural systems. Since grape cultivation is considered one of the most intensive agricultural systems [37], thus having a significant impact on biodiversity, the elaborated DSS was tested on a viticultural agroecosystem. Previous studies have tested DSSs in vineyards to support chain design and risk management assessment under sustainability goals [38]. In contrast, others have focused on identifying high-quality terroir to plan a new sustainable vineyard [39]. However, less attention has been given to the role of biodiversity in the sustainability of vineyard agricultural systems. Compared to the environmental DSS tested in vineyards and developed by Lamastra et al. [40], which included a biodiversity management indicator strictly related to landscape management, the developed DSS was conceived to provide a comprehensive and implementation-oriented assessment method.

2. Materials and Methods

2.1. Decision Support System

Indicators for assessing agrobiodiversity were selected based on the ease of obtainment, similar to other assessment methods proposed [7], and the literature review [41] used to identify relevant proposed indicators in previous studies. Table 1 displays the selected indicators and corresponding consulted references supporting their relevance.

A state-of-the-art review of biodiversity assessment methodologies concluded that the min–max scaling method is the most common method of normalisation of the indicators in the analysed studies [32]. Few methodologies aiming to assess farm biodiversity comprehensively, and not only on a genetic, population, or dietary level, use the benchmark method to aggregate indicator scores. The benchmark method confers the advantage over other approaches to normalise indicators of making methods more robust and comparable while minimising difficulty in the index’s final value interpretation [32,41]. This method considers recommended limits in the scientific literature, set by governmental restrictions, or from reputable organisations. The elaboration of the DSS partially followed this method of normalisation, as the generality of maximum and minimum values was defined according to recommended values in the scientific literature and from reputable organisations, such as the FAO, European Commission, and United States Department of Agriculture (USDA), among others. This way, the selected normalisation method quantifies the deviation of recommended sustainable values [42]. Values greater than or equal to zero correspond to sustainable performance levels and negative values correspond to values below the minimum sustainable limit [43].

It is challenging to establish a sustainable threshold for indicators like species richness of crops and average species richness of cover crops. This is because species richness depends on their tolerance/adaptability to multiple environmental factors, including water availability, soil characteristics, and management choices. Therefore, a proportion of enhancement potential defined by the user was used to calculate the maximum values for these indicators. Other threshold values can be altered according to users’ perception of adequacy, namely regarding the maximum values of buffer zone width and crop rotation. However, the maximum values displayed in Table 1 are displayed as default values in the DSS. The recommended, minimum, and maximum values are also described in Table 1.

Biodiversity is considered to have a positive contribution to natural pest control [44], yield stability, and productivity through the potential increased resistance to environmental changes [4]. However, the positive association between increased biodiversity and increased yield is not a universal condition. Previous studies have reported yield maintenance [45,46] from implementing agrobiodiversity practices in vineyards, while others verified yield reductions mainly associated with the use of cover crops [47], due to competition with the main crops [48]. To allow the DSS user to monitor whether productivity and natural pest control are positively associated with higher levels of biodiversity, the category of factors influenced by agrobiodiversity was added. The two indicators composing this category are also displayed in Table 1.

Table 1. Biodiversity indicator calculation and recommended values.

Category	Pillar	Indicator	Calculation	Recommended Values	Minimum Value	Maximum Value
Plants	Diversity of plants [49]	Species richness of crops	N—Number of cultivated species	Maximum	Minimum value of N in the analysed period.	$\mathrm{N}\times \left(1+\mathrm{p}1\right)$ where p1 is a proportion indicating enhancement potential defined by the user 1
		Average species richness of cover crops	${\displaystyle \frac{{\sum }_{\mathrm{i}=1}^{\mathrm{n}}\mathrm{S}}{\mathrm{n}}}$ Where S is the number of species in each sample and n is the sample size	Maximum	Minimum value of ${\displaystyle \frac{{\sum }_{\mathrm{i}=1}^{\mathrm{n}}\mathrm{S}}{\mathrm{n}}}$ in the analysed period.	${\displaystyle \frac{{\sum }_{\mathrm{i}=1}^{\mathrm{n}}\mathrm{S}}{\mathrm{n}}}\times \left(1+\mathrm{p}2\right)$ where p2 is a proportion indicating enhancement potential defined by the user 1
Semi-natural Habitats	Presence of natural semi-natural habitats [50]	Share of semi-natural habitats (SNH)	${\displaystyle \frac{\mathrm{SNH} \mathrm{area}}{\mathrm{Total} \mathrm{Farm} \mathrm{Area} \left(\mathrm{TFA}\right)}}\times 100$	20–25% [50]	20%	100%
Agricultural Management Practices	Buffer zone [51]	Buffer zone presence	Share of water courses in % with a buffer zone compared to total shoreline	100% [51]	99% 2	100%
		Buffer zone width	Buffer zone width compared to total shoreline (in metres)	≥10 [51]	10	50 3
	Conservation agriculture [52,53]	Crop rotation	Number of years it takes for a specific crop to be grown again in the same field [54]	≥3 years [54] or ≥6 years [55]	3	6
		Soil cover	Utilised Agricultural Area (UAA) covered by crop residues and/or cover crop (ha)/total UAA (ha)	≥30% [53]	30%	100%
		Reduced tillage	Share of agricultural area under the practice in %	100%	99% 2	100%
	Pressure of chemical mechanisms [56]	UAA not treated with inorganic fertiliser	${\displaystyle \frac{\begin{array}{c}\mathrm{UAA} \mathrm{not} \mathrm{treated} \mathrm{with}\\ \mathrm{inorganic} \mathrm{fertiliser} \left(\mathrm{ha}\right) \end{array}}{\mathrm{Total} \mathrm{UAA} \left(\mathrm{ha}\right)}}\times 100$	100%	20% 4	100%
		UAA not treated with inorganic insecticide	${\displaystyle \frac{\begin{array}{c}\text{UAA not treated with}\\ \text{inorganic insecticide} \left(\mathrm{ha}\right) \end{array}}{\mathrm{Total} \mathrm{UAA} \left(\mathrm{ha}\right)}}\times 100$	100%	50% 4	100%
		UAA not treated with inorganic fungicide	${\displaystyle \frac{\begin{array}{c}\text{UAA not treated with}\\ \text{inorganic fungicide} \left(\mathrm{ha}\right) \end{array}}{\mathrm{Total} \mathrm{UAA} \left(\mathrm{ha}\right)}}\times 100$	100%	50% 4	100%
		UAA not treated with inorganic herbicide	${\displaystyle \frac{\begin{array}{c}\text{UAA not treated with}\\ \text{inorganic herbicide} \left(\mathrm{ha}\right) \end{array}}{\mathrm{Total} \mathrm{UAA} \left(\mathrm{ha}\right)}}\times 100$	100%	50% 4	100%
Factors Influenced by Agrobiodiversity	Harvest yield [4]	Total crop production per area	${\displaystyle \frac{{\sum }_{\mathrm{i}=1}^{\mathrm{n}}\mathrm{Y}}{\mathrm{n}}}$, where n is the sample size and Y is the yield per area of the crop $Y={\displaystyle \frac{\mathrm{P}}{\mathrm{A}}}$, where P is the weight of the crop harvested, and A is the size of the cultivated area	Maximum	-	-
	Pest management [4]	Enemy-to-pest Ratio [57,58,59]	${\displaystyle \frac{\mathrm{NE}}{\mathrm{P}}}$ Where NE is the total number of captured Natural Enemies individuals and P is the total number of captured crop pests	Maximum	-	-

¹ To estimate the maximum values of these parameters, the user is asked to indicate a percentage of improvement considered reasonable, given the means at the users’ disposal and environmental factors about the minimum value (more detrimental to agrobiodiversity) in the period under analysis. ² From a sustainability viewpoint, the higher this value, the better. However, so that the minimum and maximum sustainable values are not the same (which would prevent the calculation), 99% was defined as the minimum value. ³ 50 m (165 feet) is a recommended value for creating habitats for invertebrates, aquatic species, reptiles, amphibians, birds, and mammals [51]. ⁴ From a sustainability viewpoint, the larger the area in which chemical control mechanisms are not used, the better. However, so that the minimum and maximum sustainable values are not the same (which would prevent the calculation of this indicator), the minimum values were defined by the goals defined in the “Farm to Fork” strategy by the European Union [60], which are: (1) reduce the overall use and risk of chemical pesticides by 50% (including herbicides and insecticides [61]) by at least 50% by 2030; (2) reduce the use of fertilisers by at least 20% by 2030.

Table 1. Biodiversity indicator calculation and recommended values.

Category	Pillar	Indicator	Calculation	Recommended Values	Minimum Value	Maximum Value
Plants	Diversity of plants [49]	Species richness of crops	N—Number of cultivated species	Maximum	Minimum value of N in the analysed period.	$\mathrm{N}\times \left(1+\mathrm{p}1\right)$ where p1 is a proportion indicating enhancement potential defined by the user 1
		Average species richness of cover crops	${\displaystyle \frac{{\sum }_{\mathrm{i}=1}^{\mathrm{n}}\mathrm{S}}{\mathrm{n}}}$ Where S is the number of species in each sample and n is the sample size	Maximum	Minimum value of ${\displaystyle \frac{{\sum }_{\mathrm{i}=1}^{\mathrm{n}}\mathrm{S}}{\mathrm{n}}}$ in the analysed period.	${\displaystyle \frac{{\sum }_{\mathrm{i}=1}^{\mathrm{n}}\mathrm{S}}{\mathrm{n}}}\times \left(1+\mathrm{p}2\right)$ where p2 is a proportion indicating enhancement potential defined by the user 1
Semi-natural Habitats	Presence of natural semi-natural habitats [50]	Share of semi-natural habitats (SNH)	${\displaystyle \frac{\mathrm{SNH} \mathrm{area}}{\mathrm{Total} \mathrm{Farm} \mathrm{Area} \left(\mathrm{TFA}\right)}}\times 100$	20–25% [50]	20%	100%
Agricultural Management Practices	Buffer zone [51]	Buffer zone presence	Share of water courses in % with a buffer zone compared to total shoreline	100% [51]	99% 2	100%
		Buffer zone width	Buffer zone width compared to total shoreline (in metres)	≥10 [51]	10	50 3
	Conservation agriculture [52,53]	Crop rotation	Number of years it takes for a specific crop to be grown again in the same field [54]	≥3 years [54] or ≥6 years [55]	3	6
		Soil cover	Utilised Agricultural Area (UAA) covered by crop residues and/or cover crop (ha)/total UAA (ha)	≥30% [53]	30%	100%
		Reduced tillage	Share of agricultural area under the practice in %	100%	99% 2	100%
	Pressure of chemical mechanisms [56]	UAA not treated with inorganic fertiliser	${\displaystyle \frac{\begin{array}{c}\mathrm{UAA} \mathrm{not} \mathrm{treated} \mathrm{with}\\ \mathrm{inorganic} \mathrm{fertiliser} \left(\mathrm{ha}\right) \end{array}}{\mathrm{Total} \mathrm{UAA} \left(\mathrm{ha}\right)}}\times 100$	100%	20% 4	100%
		UAA not treated with inorganic insecticide	${\displaystyle \frac{\begin{array}{c}\text{UAA not treated with}\\ \text{inorganic insecticide} \left(\mathrm{ha}\right) \end{array}}{\mathrm{Total} \mathrm{UAA} \left(\mathrm{ha}\right)}}\times 100$	100%	50% 4	100%
		UAA not treated with inorganic fungicide	${\displaystyle \frac{\begin{array}{c}\text{UAA not treated with}\\ \text{inorganic fungicide} \left(\mathrm{ha}\right) \end{array}}{\mathrm{Total} \mathrm{UAA} \left(\mathrm{ha}\right)}}\times 100$	100%	50% 4	100%
		UAA not treated with inorganic herbicide	${\displaystyle \frac{\begin{array}{c}\text{UAA not treated with}\\ \text{inorganic herbicide} \left(\mathrm{ha}\right) \end{array}}{\mathrm{Total} \mathrm{UAA} \left(\mathrm{ha}\right)}}\times 100$	100%	50% 4	100%
Factors Influenced by Agrobiodiversity	Harvest yield [4]	Total crop production per area	${\displaystyle \frac{{\sum }_{\mathrm{i}=1}^{\mathrm{n}}\mathrm{Y}}{\mathrm{n}}}$, where n is the sample size and Y is the yield per area of the crop $Y={\displaystyle \frac{\mathrm{P}}{\mathrm{A}}}$, where P is the weight of the crop harvested, and A is the size of the cultivated area	Maximum	-	-
	Pest management [4]	Enemy-to-pest Ratio [57,58,59]	${\displaystyle \frac{\mathrm{NE}}{\mathrm{P}}}$ Where NE is the total number of captured Natural Enemies individuals and P is the total number of captured crop pests	Maximum	-	-

¹ To estimate the maximum values of these parameters, the user is asked to indicate a percentage of improvement considered reasonable, given the means at the users’ disposal and environmental factors about the minimum value (more detrimental to agrobiodiversity) in the period under analysis. ² From a sustainability viewpoint, the higher this value, the better. However, so that the minimum and maximum sustainable values are not the same (which would prevent the calculation), 99% was defined as the minimum value. ³ 50 m (165 feet) is a recommended value for creating habitats for invertebrates, aquatic species, reptiles, amphibians, birds, and mammals [51]. ⁴ From a sustainability viewpoint, the larger the area in which chemical control mechanisms are not used, the better. However, so that the minimum and maximum sustainable values are not the same (which would prevent the calculation of this indicator), the minimum values were defined by the goals defined in the “Farm to Fork” strategy by the European Union [60], which are: (1) reduce the overall use and risk of chemical pesticides by 50% (including herbicides and insecticides [61]) by at least 50% by 2030; (2) reduce the use of fertilisers by at least 20% by 2030.

The minimum values correspond to 20% and 50% in the use of chemical fertilisers and pesticides, respectively.

The indicator values are normalised following Equation (1) and are aggregated according to Equation (2) [52].

$$\mathrm{Indicator} \mathrm{score} \left(\mathrm{Ii}\right)={\displaystyle \frac{\mathrm{X}-\mathrm{Lower} \mathrm{X} \mathrm{threshold}}{\mathrm{Upper} \mathrm{X} \mathrm{threshold}-\mathrm{Lower} \mathrm{X} \mathrm{threshold}}}$$

(1)

$$\mathrm{Biodiversity} \mathrm{Indicator} \mathrm{Score}= {\displaystyle \frac{\sum _{\mathrm{i}=1}\mathrm{Si}}{\mathrm{Number} \mathrm{of} \mathrm{measured} \mathrm{indicators}}}$$

(2)

For each indicator, a set of enhancement practices recommended by Marcelino et al. [11] and other reputable organisations, such as the FAO and USDA, among others, were collected, aiming to incorporate them as a DSS output. The recommended practices and corresponding bibliographic sources are displayed in

Table 2. Recommendations and corresponding benefits associated with biodiversity indicators.

Indicator	Recommendations	Benefits	Ref.
Species Richness of Crops	Cultivation of different crop varieties and, if possible, different crop species	Minimisation of the proliferation of disease and insect pests stimulated by monoculture systems	[49,62,63]
Average Species Richness of Cover Crops	Inclusion of leguminous plants into a cropping system	Nitrogen fixation in the soil, potentially benefiting overall yields Stability of income	[28,64]
	Conservation of diversity of non-crop vegetation	Pollination support Improvement of soil fertility and pest regulation Avoidance of yield losses caused by dominant weed species	[65]
Share of SNH	Conservation of a proportion of semi-natural habitat area equal to or higher than 20–25% per km2	Preservation of biodiversity regarding capacity to: Pollinate crops Regulate pests and diseases Maintain clear water Limit soil erosion	[50]
Buffer Zone Presence	Introduction of trees, shrubs, and perennials that thrive in different humidity levels adjacent to a water body	Filtering water runoff before it reaches the water body, Sediment, nutrient, and pathogen capture Soil erosion reduction. Mitigation of natural flooding and promotion of the integrity of ecosystems and habitats in water bodies	[51,66,67]
Buffer Zone Width	Ensuring a minimum width of around 10 m (35 feet) of the buffer zone consists of trees, shrubs, and perennial plants	Improvement of terrestrial and aquatic habitat Reduction in terrestrial transport of sediments and organic materials Maintenance or increase in total carbon stored in soils and/or plant biomass Reduction in atmospheric concentrations of greenhouse gases	[51]
Crop Rotation	Using at least 3 different crops and ensuring that 3 or more years pass before a specific planting is carried out in the same area of agricultural land	Improvement of soil structure Promotion of a diverse range of soil flora and fauna Better available nutrient stocks Prevention of pests and diseases	[53,54,55]
Soil Cover	Maintenance of a permanent soil organic cover of at least 30% with cover crops	Enhancement of diversity of many taxa, namely invertebrates, invertebrate, pest predators, and pollinators Soil protection from the impact of extreme weather patterns Preservation of soil moisture Minimisation of soil compaction	[53,68]
	Maintenance of permanent soil organic cover of at least 30% with crop residues	Soil organic matter increasing (depending on the type of mulch and edaphoclimatic conditions) Soil moisture conservation Soil temperature stabilisation Soil erosion reduction Soil fertility Weed suppression	[53,63,64,69]
Reduced Tillage	Minimisation of soil disturbance and retaining at least 30% of the preceding crop’s stubble on the soil surface	Mitigation of soil ecological pressure Reduction in production costs Preservation of soil quality (preventing loss of organic matter, compaction, and erosion) Biological activity	[63,70]
UAA Not Treated with Inorganic Fertiliser	Use of organic soil amendments, namely manure, compost, biochar, plant residues, among others	Improvement of physical, chemical, and biological properties of soils Provision and retaining of nutrients Climate change mitigation	[28,49]
	Avoidance of the application of inorganic fertilisers	Minimisation of soil contamination with potentially harmful elements to consumer health Minimisation of air and groundwater pollution	[42,71]
UAA Not Treated with Inorganic Insecticide	Encouraging the presence of natural enemies of pests, for instance, by providing shelter boxes for bats (1) 1	Enhancement of pollination (1) Improvement of arthropod diversity and bat activity (2) Crop loss reduction, through biological pest suppression Minimisation of damage to crops and productive gains Minimisation of soil, water, and atmospheric pollution, contamination of the food chain, and consequent diseases caused in humans and deaths of other non-target organisms (bees, fish, birds, etc.) caused by insecticide and fungicide application	[72,73,74]
	Use of repellent plants and botanical sprays instead of synthetic pesticides		[9,61]
	Sticky Traps		[75]
	Biofumigation		[75]
	Pheromones or other volatile compounds, like attractants and info chemicals		[69,75]
	Increasing landscape heterogeneity (2)		[76]
	Incorporation of companion crops, and also catch and trap crops		[75]
	Use of hydrolats and other biopesticides for ecological weed management		[28,77,78]
UAA Not Treated with Inorganic Fungicide	Plantation of fungus-resistant varieties 2	Augmentation of the abundance of arthropods in general and beneficial arthropods in particular Minimisation of soil and water contamination Maintenance of biological soil activity	[79,80,81]
UAA Not Treated with Inorganic Herbicide	Implementation of preventive measures, namely: Using clean seeds Keeping the seed bed free from weeds Using well-decomposed organic manures Keeping the bunds and irrigation channels free from weeds Keeping tools and farm machinery clean Controlling weeds before they attain the reproductive stage	Reduction in herbicide use, minimising the mortality caused to non-target organisms (worms, fish, etc.), the decline in pollinators, deterioration of groundwater quality, and other hazards caused by herbicide application Minimisation of soil and water contamination Enhancement of nutrient availability (1)	[64]
	Solarisation (1)		[64]
	Use of mulches		[64]
	Use of mechanical tools such as torsion weeders, finger weeders, etc.		[64]

¹ Recommendations (1) and (2) are associated specifically with benefits (1) and (2), respectively, while the remaining benefits are associated with all the remaining recommendations. ² Recommendations for the indicator “UAA not treated with insecticides” are also applicable to the indicator “UAA not treated with fungicide”. However, to avoid repetition, only the recommendation “Plantation of fungus-resistant varieties” was included.

.

Following the described methodology, the DSS was designed using Microsoft^® Excel version 2209, due to its ease of use and reduced implementation and maintenance costs, when compared to other computerised solutions [82].

2.2. Data Collection

The developed DSS was tested on a commercial vineyard (Colinas do Douro) located in the Douro Superior wine-growing region, Portugal (40.99° N; 6.98° W) at an elevation of 463 to 477 m and a slope of 9%. The required data for testing the DSS were mainly collected by field surveys and farmers’ interviews, following the questions integrating the DSS, which are described in the next section.

To determine cover crop species richness, in April (the period of maximum flora richness), all taxa were identified inside 16 randomised 0.5 m² per interrow (4 plots of 4 samples) areas of an organic plot. The average number of plant species per m² was used to calculate the specific richness indicator. The indicator “Average species richness of cover crops”, mentioned in Table 1, was calculated as the average of the 16 samples.

To assess grape yield, 40 sentinel vines (4 randomised sets of 10 contiguous vines each) were selected and the number of clusters and grape weight per vine were assessed at harvest. The average grape yield per ha was used to calculate the harvest yield indicator.

As no data were available based on captures of natural enemies, we considered a proxy calculated as the ratio between the counting of Hymenoptera (many are parasitoids or predators) and those of the European grape moth, Lobesia botrana, in automatic sex pheromone traps used for monitoring this grapevine pest, between April/May and August. Although these traps are aimed at selectively attracting and capturing males of the target insect pest, they also capture other insects intercepted by the sticky plates used within the trap.

3. Results

This section is intended to present the DSS testing with the data collected on a viticulture farm located in the Douro Region to demonstrate the reproducibility and utility of the elaborated DSS.

The DSS is composed of four sections. In the first one, the “Data” section, the user provides the needed information to visualise biodiversity performance assessment. The DSS was configured in a way to allow the user to select which answers apply to the farm context, providing feedback on the selected topics by the user.

The DSS was initially tested with data from two years to validate its functionality and effectiveness. Despite this, the system is designed to accommodate and analyze data over extended periods, providing a more comprehensive view of agrobiodiversity performance. This flexibility ensures that users can tailor the DSS application to the length of their available datasets and specific analytical needs. Consequently, while the present study demonstrates the DSS using two years of data, its design inherently supports longer-term analyses, enhancing its utility for sustainable agricultural management. Since the DSS was developed using Microsoft^® Excel, the user can obtain biodiversity performance assessment and decision support for the desired number of years by simply adding new columns and dragging the fill handle to copy formulas.

Table 3. Questions and answers for the biodiversity indicator calculation of the Decision Support System (DSS).

Number of Question	Questions	Inclusion	Year
			2022	2023
1	What is the average total number of cover crop species per square metre that you can identify in the agricultural area?	Yes	21.0	34.0
2	Indicate the percentage of improvement in cover crop diversity that you consider viable on the analysed agricultural land.	N.A.	10%
3	What is the total number of cultivated plant species that you have on the analysed agricultural land?	Yes	2	2
4	Indicate the percentage improvement in the diversity of cultivated plants that you consider viable on the analysed agricultural land.	N.A.	1%
5	What is the total farm area (in ha)?	N.A.	476.0	476.0
6	What is the total utilised agricultural area of the farm (in ha)?	N.A.	116.0	116.0
7	Which area is covered with semi-natural habitats (in ha)?	Yes	360.0	360.0
8	If you have any water body on your farm, what percentage of the total shoreline has a buffer zone?	Yes	25%	25%
9	If you have any water body on your farm, on average, what is the width (in m) of the buffer zone created along the total shoreline?	Yes	5.0	5.0
10	If you have annual or biannual crops, indicate the time interval (in years) that passes until the same crop is planted again in the same area of land.	No
11	What is the utilised agricultural area of the farm with soil cover (in ha)?	Yes	5.5	5.5
12	What is the utilised agricultural area of the farm with reduced tillage practice (in ha)?	Yes	5.5	5.5
13	What is the utilised agricultural area of the farm not treated with inorganic fertilisers (in ha)?	Yes	30.5	30.5
14	What is the utilised agricultural area of the farm not treated with inorganic insecticides (in ha)?	Yes	116.0	110.0
15	What is the utilised agricultural area of the farm not treated with inorganic fungicides (in ha)?	Yes	116.0	116.0
16	What is the utilised agricultural area of the farm not treated with inorganic herbicides (in ha)?	Yes	26.0	26.0
17	What yield (number of kg per ha) were you able to obtain from your harvest?	Yes	4992.7	11,491.8
18	If you have traps to monitor major crop pests, please indicate the number of captures throughout the year for one of those pests.	Yes	190.0	32.0
19	If you have traps to monitor natural enemies (e.g., yellow sticky traps), please indicate the number of natural enemies’ captured throughout the year; if not, please indicate the number of natural enemies captured in the traps mentioned in Question 18.	Yes	15	55

exhibits the answers obtained with collected data on the viticultural farm for the two studied years.

From the analysed viticultural farm, it was only possible to collect data for two years. It can be observed that the answer to Question 10, related to crop rotation, was excluded from the analysis. It was verified that the number of cover crop species increased in 2023 as well as the harvest yield. In the same year, the utilised agricultural area that was not treated with inorganic insecticides and the number of captures of crop pests decreased, compared to 2022.

Based on the obtained answers, the DSS displays in the following section the evaluation obtained for each indicator, presenting a colour code. If the value is greater than or equal to zero, the cell is coloured green. An indicator with an assessment value of zero means that it meets the minimum sustainable value, and a positive value means that the indicator score exceeds the minimum sustainable value. On the other hand, cells coloured in red represent indicator scores with values lower than the minimum sustainable value. The evaluation of “biodiversity performance” provided by the DSS is displayed in

Table 4. Evaluation of Biodiversity Performance provided by the decision support system.

Category	Indicator	Inclusion	Minimum Value	Maximum Value	Real Value		Evaluation
					2022	2023	2022	2023
Plants	Average species richness of cover crops	1	21	23.1	21	34	0.00	6.19
	Species richness of crops	1	2	2.01	2	2	0.00	0.00
Semi-natural Habitats	Percentage of semi-natural habitats	1	20%	100%	76%	76%	0.70	0.70
Agricultural Management Practices	Buffer zone presence	1	99%	100%	25%	25%	−74.00	−74.00
	Buffer zone width (m)	1	10	50	5	5	−0.13	−0.13
	Crop rotation	0	3	6	Excluded	Excluded	Excluded	Excluded
	Percentage of soil cover	1	30%	100%	5%	5%	−0.36	−0.36
	Percentage of reduced tillage implementation	1	99%	100%	5%	5%	−94.26	−94.26
	Percentage of utilised agricultural area not treated with inorganic fertilisers	1	20%	100%	26%	26%	0.08	0.08
	Percentage of utilised agricultural area not treated with inorganic insecticides	1	50%	100%	100%	95%	1.00	0.90
	Percentage of utilised agricultural area not treated with inorganic fungicides	1	50%	100%	100%	100%	1.00	1.00
	Percentage of utilised agricultural rea not treated with inorganic herbicides	1	50%	100%	22%	22%	−0.55	−0.55
Number of Indicators Included in the Analysis		11					−15.14	−14.59

.

As can be seen in Table 4, the improvement in the biodiversity performance verified in 2023 compared to 2022 was due to the higher average species richness of cover crops. This increase in cover crop richness may have contributed to the reduction in the percentage of utilised agricultural area not treated with inorganic insecticides. The results exhibited in this DSS section are consistent with the answers obtained in the previous section, according to Table 3.

The next section presents a list of sustainable practices (corresponding to those presented in Table 2) displayed in order of priority, as the data are ordered from the lowest to the highest assessment value for each indicator, calculated in the “biodiversity performance” section. Since the indicator percentage of reduced tillage implementation has the highest pejorative value in terms of recommendations for biodiversity enhancement in 2023, a list of sustainable practices and corresponding benefits are displayed in the first place, followed by the remaining indicators with better performance.

Table 5. Recommendations for sustainable practices and expected benefits provided by the Decision Support System (DSS).

Category	Indicator	Recommended Value	Evaluation	Sustainable Practices	Benefits
Agricultural Management Practices	Percentage of reduced tillage implementation	99%	−94.26	Minimisation of soil disturbance and retaining at least 30% of the preceding crop’s stubble on the soil surface.	Mitigation of soil ecological pressure Reduction in production costs Preservation of soil quality (preventing loss of organic matter, compaction, and erosion) Biological activity
Agricultural Management Practices	Buffer Zone Presence	99%	−74.00	Introduction of trees, shrubs and perennials that thrive in different humidity levels adjacent to a water body.	Filtering water runoff before it reaches the water body, Sediment, nutrient, and pathogen capture Soil erosion reduction. Mitigation of natural flooding and promotion of the integrity of ecosystems and habitats in water bodies

shows the sustainable practices and associated benefits of the first two indicators presented in the DSS.

Lastly, in the “Impact Monitoring” sheet (Figure 1), the user can visualise the temporal evolution of the harvest yield and the average number of captures of natural enemies (pests) per sample in the form of charts to allow comparison with the evolution of biodiversity performance. The table of Figure 1a displays the indicator values and the table of Figure 1b presents the corresponding percentage variation.

Between 2022 and 2023, a 4% increase in the biodiversity indicator was verified. During the same period, there was an increase of 130% in the harvest yield and an increase of 2077% in the enemy-to-pest indicator, which can be associated with better control of pests. The exhibition of this information allows the visualisation of the evolution of harvest yield and pest control (associated with the enemy-to-pest indicator) along with the biodiversity indicator, as recommended practices are implemented throughout time. This information can help users identify the need to adapt vineyard management practices to obtain desired levels of harvest yield and pest control while promoting biodiversity.

4. Discussion

Biodiversity is a broad concept and is hard to quantify in its plenitude. However, the use of appropriate assessment methods can be a tool to monitor biodiversity impact over time and identify opportunities for improvement.

The DSS presented in this study can enable decision makers in the agricultural sector to contribute to the achievement of SDG 15 “Life on land”, envisioning to halt biodiversity loss and reverse land degradation [13]. The elaborated DSS distinguishes itself from previously developed assessment methods since it provides a list of sustainable practices in order of priority, acting as an action plan that converts quantitative assessment values into practical suggestions.

In summary, the developed DSS allows the users to do the following:

• Analyse agrobiodiversity performance over 5 years or longer.
• Select which data the user wants to include in the analysis.
• Define thresholds to assess some of the indicators according to their perception.
• Visualise the assessment with a colour code for each biodiversity indicator.
• Visualise a set of sustainable practices in order of priority and ease of implementation.
• Monitor the temporal evolution of factors that are theoretically positively influenced by higher levels of biodiversity performance, namely harvest yield and pest control.

The analysis of the Impact Monitoring Section tested with data from a viticultural farm showed an improvement of 4% in the biodiversity indicator. In the same period, there was an increase of 130% in the harvest yield. However, in the year 2022, the amount of precipitation and water availability were notably lower than in 2023. This factor significantly influenced the harvest yield increase. In the same period, there was an increase of 130% in harvest yield and of 2077% in the indicator related to pest control. The increase in this last indicator was a result of a decrease in the population level of L. botrana and an increase in Hymenoptera captures. The higher activity of Hymenoptera in 2023 may be explained, at least in part, by the observed increase (29%) in species diversity of cover crops. Cover crops have been shown to enhance beneficial arthropod communities, including Hymenoptera [83]. The study by Raffa et al. [45], conducted from 2017 to 2020, has associated the maintenance of cover crops with increased grape yield. However, the present study does not allow associating the increase in species richness of cover crops with the increase in yield since it only analysed two years. The improvement in the biodiversity indicator (4%), due to the increase in the average species richness of cover crops, as exhibited in Section 3, as well as the crop yield, were mainly influenced by higher water availability in 2023 as compared to 2022. Despite an improvement of the biodiversity indicator being verified along with an increase in the enemy-to-pest ratio and an increase in yield, users should be aware that such variations are not necessarily explained by the improvement of biodiversity performance, as quantified in the developed DSS. A wide variety of factors not contemplated in the DSS can influence both pest control and yield. The data displayed in the Impact Monitoring Section do not confirm the existence of a correlation between the indicators. Instead, the data only allow a visual comparison. Further investigation is needed to assess the impact of biodiversity enhancement on yield variation and pest control in the long term.

The decision support system (DSS) was tested using two years of data collected in the field to demonstrate its functionality and potential for assessing agrobiodiversity performance. However, the DSS can analyze biodiversity performance over longer periods. A longer time frame enables a more robust and comprehensive evaluation of trends and variations in agrobiodiversity. While shorter periods can also be used to capture initial trends, longer periods will provide a more balanced approach to observing significant changes and patterns, which may be necessary for more accurate and meaningful assessments. Conversely, extending the analysis to a longer period, while potentially yielding even deeper insights, may also introduce additional complexities and data collection challenges. The flexibility of the DSS design allows it to adapt to various time spans according to data availability and specific requirements of the users. Therefore, while this study utilised two years of data for initial testing, the system’s capabilities extend to multi-year analyses, supporting more refined decision making in agricultural management.

Although previous studies have presented the DSS as encouraging sustainable farming [10,84,85,86], the here-described DSS makes an innovative contribution since, to our knowledge, it is the first DSS for biodiversity assessment in a farm context that incorporates recommendations and associated benefits, considering updated recommendations of sustainable practices provided by the recent literature, the FAO and European guidelines. Despite making an innovative contribution concerning biodiversity assessment methods, this DSS is subject to other limitations, namely the following:

• A correlation test between the selected indicators was not performed and it was therefore not possible to demonstrate the absence of redundancies between them.
• The arithmetic mean aggregation method applied leads to compensation effects since lower values of an indicator are compensated by higher values of another [31]. However, the loss of information potentially associated with this aggregation method is minimised through the action plan exhibition.
• The biodiversity indicator does not capture the specific and functional diversity of soil microorganisms nor the diversity of animals and the selected assessment method’s applicability to livestock farms can be reductive.
• The presented recommended practices for biodiversity may not be economically, environmentally, and socially sustainable in every context. Sustainable development in the agricultural sector is considered a continuous process to reach a balance between economic, social, and environmental benefits [87]. Hence, the recommended practices’ implementation should be adapted to users’ perception of adequacy.

To overcome these limitations, further research is recommended regarding the following:

• The development of the DSS with the possibility to record lessons learned by the user about the viability of the proposed practices for biodiversity enhancement to adapt the recommendations presented in future situations. This new functionality could be added to an app version of the DSS to be developed.
• The inclusion of users’ perception of local constraints that, beyond variation in agrobiodiversity performance, may influence crop yield and pest control.
• Evaluation of the DSS’s efficacy in promoting on-farm biodiversity enhancement in the long term.

5. Conclusions

The degradation of natural ecosystems, due to agricultural intensification and overexploitation of natural resources has been a key driver of the unprecedented biodiversity loss in the last 50 years. Appropriate implementation-oriented assessment methods can empower decision makers on farms to make their contribution to halting biodiversity loss. This paper described the methodology followed and demonstrated that the elaborated DSS correctly calculates the biodiversity indicator and communicates through a colour code regarding the areas that need improvement. Despite being tested on a viticultural farm, the DSS can be applied to stockless farms of diverse agricultural subsectors, since it incorporates generic recommendations for agrobiodiversity enhancement.