The impact of intercropping, tillage and fertilizer type on soil and crop yield in fruit orchards under Mediterranean conditions: A meta-analysis of field studies

Literature searches were conducted in SCOPUS, targeting peer reviewed articles published online through July 2019. The results included 187 experimental treatments from 46 peer-reviewed articles, including the countries of Spain, Italy, France, Portugal, Greece, Turkey, Slovenia, Tunisia, Chile and the United States of America. To obtain the papers listed, we used the keywords “almond OR olive OR citrus OR vineyard OR grapevine OR orchard”, “AND” combinations of the following items: “Mediterranean climate”, “cover crop”, “intercropping”, “alley crop”, “multicrop”, “agroforestry”, “crop yield” and “soil”. Laboratory or greenhouse studies were excluded, and only studies performed in field conditions and carried out in a Mediterranean climate were chosen. When several papers included data from the same experiment, the longest study with the most soil and crop variables was chosen.

A database was compiled with the relevant data extracted for (i) management practices (crop diversification, tillage, fertilization); (ii) environmental characteristics, such as soil clay and climatic variables, (iii) study length and (iv) response variables (total nitrogen (N); available phosphorus (P), soil organic carbon (SOC) and crop yield). Very few articles showed the yield of the alley crops, so we were unable to perform a meta-analysis on alley crop yields (n < 8) (Lee et al., 2019). Consequently, only the tree crop yield is used in this meta-analysis. The database is included as an Excel file in the Supplementary Material, including soil depth used and specific soil characteristics and management. The fruit tree crops used for the study were mostly grapevines (Vitis vinifera L.) at 36% of the sample size, olive trees (Olea europaea L.) at 34% of the sample size, almond trees (Prunus dulcis (Mill.) D.A. Webb) at 15% of the sample size and citrus trees (Citrus x sinensis Osbeck, Citrus x limon (L.) Osbeck) at 7% of the sample size. We also used other fruit trees, such as avocado (Persea americana Mill.), carob (Ceratonia siliqua L.), peach (Prunus persica (L.) Stokes), chestnut (Castanea sativa Mill.) and walnut (Juglans regia L.), representing 8% of the total dataset.

To assess the effects of intercropping, tillage and fertilization on soil N, soil P, SOC and yield, we individually studied the following categories (percentages between brackets indicate the frequency of each subgroup within each category):

• A
  Diversification type, comparing permanent intercropping (PC) (45%) and annual intercropping (AC) (55%) against mono-cropping. Permanent intercropping refers to the maintenance of a permanent cover crop in the alleys, such as aromatics (Thymus sp, Lavandula sp, Salvia sp, Rosmarinus sp, Brachypodium sp, Asparagus sp or natural grass), while annual intercropping means the presence of cover crops in the alleys that are annually harvested or incorporated into the soil. Mono-cropping indicates the presence of the tree crop alone with no other vegetation cover in the alleys (bare soil).
• B
  Tillage intensity, comparing conservation tillage (no-tillage (NT) (66%) and minimum tillage (MT) (34%)) against conventional ploughing. Minimum tillage involves residue retention in soil since cover crops or weeds are incorporated into the soil.
• C
  Fertilization type, comparing organic fertilizer (Org) against conventional inorganic fertilizer. Organic fertilizers contain plant or animal-based materials that are either a by-product or an end product of naturally occurring processes, such as humic acids, manure, compost, sludge or crop residues. Inorganic fertilizers are artificially manufactured and contain minerals or synthetic chemicals.

The influence of crop diversification, tillage and fertilization on soil and crop characteristics was studied through meta-analysis. The response ratio was calculated as the effect size unit for each indicator (Aguilera et al., 2013a). This ratio has been widely used for meta-analysis in agricultural studies (e.g., Aguilera et al., 2013a; Lee et al., 2019; Li et al., 2018). The response ratio (RR) was calculated as a proportionate change in the indicator value of the treatment group (sustainable management: intercropping, organic fertilization and no-tillage/minimum tillage) ($\overline{X}$ _SM) compared to the pairwise control group (conventional management: mono-cropping, inorganic fertilization and conventional tillage) (($\overline{X}$ _C)) (Eq. (1)):(1)$RR=\frac{{\overline{X}}_{SM}}{{\overline{X}}_C}$

The natural logarithm of the response ratio was used for the analysis (log (RR)) (Eq. (2)). This transformation linearizes the metric in small samples (Hedges et al., 1999).(2)$\text{L}\text{o}\text{g}\left.\text{R}\text{R}\right)=\text{ln}\frac{{\overline{X}}_{SM}}{{\overline{X}}_C}=\text{ln}\left.{\overline{X}}_{SM}\right)–\text{l}\text{n}({\overline{X}}_C)$

Positive values indicate higher values in the sustainable management, while negative values indicate higher values in the conventional management. In meta-analysis, studies are usually weighted by the inverse of their variance, but this information was not provided in many of the selected studies. We decided to weight observations using the sample sizes to ensure statistical significance by keeping a sufficient sample size: studies with larger sample sizes were weighted higher during aggregation (Adams et al., 1997). The weighted log (RR) (Wlog(RR)) of management option i was calculated as Eq. (3):(3)$\text{W}\text{l}\text{o}\text{g}{(\text{R}\text{R})}_{\text{i}}=\frac{1}{N_i}\sum {(\text{log}(RR)}_{ij}x{W}_{ij})$where N_i is the number of studies for the management i, log(RR)_ij is the log(RR) of management i in study j, and W_ij is the weight, calculated as Eq. (4):(4)${\text{W}}_{\text{i}\text{j}}=\frac{N_i^{SM}{N}_i^C}{N_i^{SM}+N_i^C}$where $N_i^{SM}$ and $N_i^C$ are the sample sizes of the sustainable and conventional management of the management i in study j, respectively (Adams et al., 1997). For the uncertainty analysis, we reported the means and 95% confidence intervals (CIs) of the Wlog(RR). The CIs were constructed by non-parametric bootstrapping (nboot = 10 000). The bootstrapping was only carried out for treatments with n ≥ 8, as the bootstrap is unreliable when the sample size is too small (Lee et al., 2019). We considered the effect of treatment as significant if the 95% bootstrap CI did not overlap with zero (Lee et al., 2019).

C sequestration rate was calculated according to Eq. (5):(5)$\text{C}\text{s}\text{e}\text{q}\text{u}\text{e}\text{s}\text{t}\text{r}\text{a}\text{t}\text{i}\text{o}\text{n}\text{r}\text{a}\text{t}\text{e}=\frac{(C_t-C_{t^{\text{'}}})}{t}$where C_t and C_t’ are SOC stocks (Mg C ha⁻¹) at the end and at the beginning of the experiment, respectively, and t is the duration of the experiment (years). In most cases, data on initial SOC content were not available, and so they were equalled to SOC stocks in the conventional treatment. This is justified because only comparisons performed under similar soil and climatic conditions were selected for the analysis, and we focused only on the relative performance of sustainable management versus conventional management (Aguilera et al., 2013a). Most articles did not show SOC stocks, so they were calculated as indicated in equation 6:SOC stock = d ρ SOCwhere d and ρ are soil depth (m) and bulk density (Mg m⁻³) in the soil Ap horizon, respectively. For those studies not showing d or ρ, we could not calculate the SOC stock. We could not assess the effect of fertilization on C sequestration rate since very few studies dealing with organic fertilizers and SOC included the bulk density.

Correlations and regressions between Wlog(RR) of SOC, N, P and yield and the variables soil clay, mean annual temperature, mean annual precipitation and study length were carried out to assess the effect of these environmental variables and the length of study on the response rate of the analysed properties. Statistical analyses were performed with the software IBM SPSS Statistics 24.

Both MT and NT were associated with a significant increase in SOC compared to conventional tillage (Fig. 1a). There was high overlapping in the response ratio between MT and NT, with Wlog(RR) = 0.28 and 0.26 for MT and NT, respectively. On average, C sequestration rate also increased by 1.54 mg ha⁻¹ yr⁻¹ in MT and 1.40 mg ha⁻¹ yr^-1 in NT, with 75% and 78% of studies showing a positive rate, respectively (Fig. 1b). MT significantly increased soil N compared to conventional tillage (Wlog(RR) = 0.21). However, NT systems had no significant negative effect on N compared to conventional tillage, with Wlog(RR) = 0.09 (Fig. 2a), although 71% of the studies showed a positive effect. Tillage systems had no significant effect on soil P (Fig. 2b) or crop yield (Fig. 3). However, 78% and 36% of studies showed a positive effect on soil P for MT and NT, respectively.

Organic fertilization was associated with a significant increase in SOC and N compared to inorganic fertilization (Figs. 1a and 2 a). This increase was higher for SOC (Wlog(RR) = 0.20) than for N (Wlog(RR) = 0.14). The use of organic fertilizers did not significantly affect soil P, with 50% of studies showing a positive effect (Fig. 2), or yield, with 67% of studies showing a positive effect (Fig. 3) compared to inorganic fertilizers. These results indicate that most studies showed positive effects of organic fertilization compared to inorganic fertilization.

Mediterranean orchards can provide many ecosystem services, such as food production, biodiversity or climate change mitigation (Aguilera et al., 2013b; Lee et al., 2019). However, they are influenced by socioeconomic factors, such as subsidies, farm mechanization, permanent accessibility to fertilizers, government policies, international agreements and farmers’ training, that have led overall to intensive mono-cropping systems. This cropping system may compromise long-term sustainability owing to soil degradation and loss by intensive tillage and maintenance of bare soil in the alleys during the entire growing season (Almagro et al., 2016; Almagro and Martínez-Mena, 2014; Parras-Alcántara et al., 2016). The overall absence of cover crops and the intense tillage in Mediterranean orchards is related to farmers’ perceptions about competition with the cash crop for nutrients and especially water, and the traditional belief that a “clean” and “tidy” orchard must always be free of vegetation except for the trees (Cerdà et al., 2018, 2017; Ramos et al., 2010). This perception is magnified during the pronounced dry season, a characteristic of the Mediterranean climate, which causes long periods of water stress due to potentially higher levels of evapotranspiration than precipitation in many areas (Almagro and Martínez-Mena, 2014; Guzmán et al., 2019; Pérez-Álvarez et al., 2015; Sastre et al., 2018; Zornoza et al., 2018). However, the combination of soil degradation and climate change can increase the vulnerability of the Mediterranean orchards and therefore their long-term sustainability, compromising the economy depending on them (Dono et al., 2016; Rocamora-Montiel et al., 2014). As a consequence, actions are needed to revert the current land degradation trends through adoption of sustainable cropping systems that decrease soil loss by erosion, increase soil fertility and biodiversity and secure food production by maintenance of high yields.

Crop diversification and conservation agriculture is now increasingly recognized for its capacity to minimize trade-offs between ecosystem services and maximize synergies between them (Li et al., 2019; Rosa-Schleich et al., 2019; Yang et al., 2019). This meta-analysis helps confirm the overall benefits of alley cropping in orchards that associated with conservation tillage, organic fertilization on soil quality and health, and C sequestration in Mediterranean agroecosystems due to the maintained inputs of SOM, which improve soil structure and fertility and are therefore associated with increased soil biological activity (Álvarez et al., 2007; Bastida et al., 2012; Castro et al., 2008; Ramos et al., 2010; Vignozzi et al., 2019). Although the discussion continues regarding the most appropriate soil properties or approaches able to indicate soil quality or soil health, SOC and N are two of the most commonly used internationally recognised soil quality indicators owing to the ecosystem services they provide (Zornoza et al., 2015). Therefore, the increases observed in the SOC, soil N and C sequestration rates in this study following the adoption of sustainable cropping systems, associated with high yields, can be used as indicators of soil quality improvement with maintenance of land productivity. This finding aligns with the principles addressed by Letey et al. (2003) to assess soil quality, focusing more on soil use, not function or capacity, as the criteria for attribute evaluation. These results, therefore, should be used to raise farmers’ awareness about the strengths of the cropping practices assessed through this meta-analysis, which achieve the combined goals of high crop production, low environmental degradation and sustained use of resources (Letey et al., 2003; Zornoza et al., 2015).

The use of intercropping, associated with conservation tillage, is characterized by minimal soil disturbance, ground cover in the alleys during at least some periods of the crop cycle, improved biodiversity and biological processes in the soil, and the preservation of multiple ecosystem services (Aguilera et al., 2013a; Chamizo et al., 2017; Lee et al., 2019; Taguas et al., 2015). Annual or permanent crops can be grown in the alleys solely to prevent soil erosion and enhance soil quality, water retention, soil nutrients, contribute to weed suppression and attract pollinators and auxiliary fauna, which are especially important effects in desertification-prone Mediterranean agroecosystems (e.g., Almagro et al., 2016; Chamizo et al., 2017; Hernández et al., 2005; Koch et al., 2015; Ramos et al., 2011; Taguas et al., 2015). However, they can also be grown to provide, despite the latter ecosystem services, an alternative crop with economic profitability for the farmer, such as food, feed or industrial products, increasing the resilience of the system to pest/disease incidences and market volatility (e.g., Barnes et al., 2015; Bateni et al., 2019; Cardinael et al., 2015; Hugo et al., 2008; Querné et al., 2017). Most alley crops used in Mediterranean orchards are either natural grasses, cereals, legumes or a mixture of legumes and cereals for environmental benefits or aromatics, such as lavender, sage, rosemary, thyme, alfalfa, fava bean, wheat, barley or asparagus for direct economic profitability (Almagro et al., 2016; Bateni et al., 2019; Cardinael et al., 2015; Chamizo et al., 2017; Hugo et al., 2008; Querné et al., 2017; Ramos et al., 2011, 2010; Taguas et al., 2015). Independent of the final objective of the alley crops, they were shown to have increased SOC and C sequestration rates with permanent or annual intercropping and soil N content in permanent intercropping, contributing to the maintenance of high quality soils.

In our meta-analysis, SOC and N were positively influenced by the presence of alley crops, independent of the orchard type. The presence of permanent vegetation had a more positive effect than the growth of annual crops, mostly for N. Permanent vegetation can enhance SOC and N through the continuous release of root exudates and litter deposition content, contributing to effective C sequestration (Sastre et al., 2018). The growth of annual crops is normally related to bare soil periods with at least minimum tillage. Therefore, annual alley crops have lower potential to increase SOM as much as permanent vegetation, related to the lower C inputs to the soil (Belmonte et al., 2018; Daryanto et al., 2018). When permanent cover crops were used, the P content in the soil did not follow the increasing trend observed with N, which may likely be due to the higher P demands of the permanent vegetation growing in the alleys.

The synergy between intercropping and conservation tillage is associated with physical soil improvement, water conservation and the prevention of increases in CO₂ emissions (Almagro et al., 2016; Simoes et al., 2014). Intensive tillage is applied by farmers to promote soil aeration and water infiltration and to reduce the growth of weeds. However, it can have other drawbacks, such as the destruction of soil aggregates, decreases in soil aggregate stability, damage to soil organisms and an increase in soil vulnerability to erosion (García-Díaz et al., 2018; Gómez et al., 2009; Ruiz-Colmenero et al., 2013). In addition, intensive tillage may induce oxidation of organic compounds by aerobic microorganisms due to increased aeration, which could result in a loss of SOC and increased greenhouse gas emissions (Palese et al., 2014a, 2014b). Conservation tillage, however, is usually associated with increases in SOC concentrations by decreasing SOM mineralization, which is associated with increases in soil water retention (Almagro et al., 2016; Garcia-Franco et al., 2015). In addition to decreasing aeration and disruption of soil aggregates, this positive effect on soil organic matter due to conservation tillage is also related to the stratification of SOM in different depths, which has been reported to positively affect erosion control and nutrient and water conservation (Hernanz et al., 2009). Despite the absence of differences with regard to NT and MT in SOC, N showed higher values in MT than in NT. This finding can be related to the high soil compaction and sealing that takes place in dry Mediterranean areas when NT is applied as a consequence of the low SOM content, leading to reduced N mineralization rates and the resulting low soil N availability (López-Garrido et al., 2014; Martínez-Mena et al., 2013). Nonetheless, the average value of soil depth in the studies used in this meta-analysis is 19 cm, ranging from 5 to 60 cm (see the database attached as Supplementary Material). The effect of conservation tillage on deeper soil layers, therefore, could not be assessed within this meta-analysis.

The use of organic fertilizers has been proposed to enhance soil quality and carbon sequestration in comparison to inorganic fertilizers (Montes-Borrego et al., 2013; Parras-Alcántara et al., 2015). Moreover, the use of lignin-rich organic fertilizers in fruit orchards can create an organic horizon with high biological activity, which could transfer the organic C to the mineral horizons (López et al., 2014). Our results demonstrate the increase of SOC in orchards with organic fertilizer application compared to inorganic fertilizers. In addition, from a climate change mitigation approach, the addition of organic fertilizers is also related to the reduction in the use of synthetic fertilizers and the GHG emissions related to their production (Aguilera et al., 2013b). However, the growth of permanent crops and conservation tillage resulted in a larger increase in SOC than the addition of organic fertilizers (Fig. 1). Thus, although the replacement of inorganic fertilizers by organic fertilizers is a positive strategy to increase soil organic matter, the adoption of diversified cropping systems with conservation tillage is even more effective. This disparity may be related to the protection exerted by vegetation on the soil, with the release of root exudates, accumulation of litter, and incorporation of crop residues into the soil as green manure in the case of annual crops.

It is interesting to highlight that the response ratios of the different soil properties analysed were not correlated with the climatic variables of mean annual temperature and mean annual precipitation. Thus, these results suggest that the response of SOC, soil N and soil P to management practices is not highly dependent on the specific climatic conditions of the region, with positive effects of intercropping, conservation tillage and organic fertilization on soil quality occurring independent of the specific heat or aridity of the area. Although no significant effect of the experimental length of study was observed in our analysis, this lack of effect could have been influenced by the relatively small quantity of papers showing long-term results. Only 40% of the experimental treatments had an experimental length > 5 years, and only 13% of them had an experimental length > 10 years. Thus, there is a need for publication of studies dealing with the long-term effect of intercropping, conservation tillage and organic fertilization in Mediterranean orchards to assess their trends over time.

Food production is the primary function of orchards, but unsustainable cropping systems and climate change threaten high crop yields in Mediterranean climates (Iglesias et al., 2011). Our results showed a potential for alternative cropping systems with no negative effects on the cash crop yield, but with benefits for soil quality and biodiversity. Despite farmers’ perceptions about reduced crop yield with the implementation of alley crops, conservation tillage or organic fertilizers, these practices have not provided negative effects on crop yields compared to conventional systems. This result supports that alley crops, no-tillage or reduced tillage and the incorporation of nutrients as organic amendments affect soil structure and fertility positively, with higher availability of nutrients and water retention due to increased SOM and soil aggregation (Almagro et al., 2017; Daryanto et al., 2018; Montanaro et al., 2017). Moreover, the presence of a vegetation cover in the alleys of orchards can provide habitat for arthropod predators and parasitoids, which promote biological control by feeding on pests and microorganisms (Paredes et al., 2015).

Alley crops showed no negative overall effect size on crop yield. However, it is true that some studies used in this meta-analysis detected that the growth of annual crops and mostly permanent crops can have negative effects on the tree yield (e.g., Atucha et al., 2013; Hernández et al., 2005; Klodd et al., 2016; Ripoche et al., 2011; Taguas et al., 2015). Furthermore, while climatic conditions were not correlated to the response ratio of soil properties to the analysed management practices, they were related to the response ratio of the crop yield. Our results showed that the response ratio of crop yields tended to be lower in those areas with higher temperature and lower precipitation compared to the control (Fig. 4), and therefore, there were lower increases or even decreases in crop yields compared to the conventional systems. Therefore, farmers’ perception that alley crops can decrease soil moisture and consequently compete with tree crops for water seems to be important in those areas with higher aridity and warm temperatures. As a consequence, to avoid the excessive competition with alley crops which can negatively affect the cash crop yield, the use of annual crops can be used instead of permanent crops that need to be efficiently managed. The growth of annual crops allows the management of the ground cover to remove the secondary crops in the critical periods when competition for water or nutrients is high, such as summer or when the fruit is forming, avoiding negative effects on yields (Daryanto et al., 2018; Palese et al., 2014a, 2014b). The nutrient requirements of the trees are higher when the fruit is forming, so the early elimination of competition for nutrients by alley crop removal at this critical stage should be encouraged (Palese et al., 2014a, 2014b). Furthermore, it has been reported that due to their deep and fibrous rooting systems, species from the Poaceae family generally result in higher evapotranspiration rates than legumes (Duval et al., 2016), so legumes could be good candidates for alley cropping in dry areas. According to our results, a limit of 470 mm of mean annual precipitation and a mean annual temperature of 15.5 °C may be suggested as thresholds for which caution should be exercised by farmers to avoid negative effects on tree crop yields (Fig. 4). A thorough management of the alley crops in regions with temperatures over and precipitation under this threshold should be implemented with the removal of annual crops during those critical periods when there can be competition with the cash crop, as discussed above.

Conventional tillage is usually performed to remove weeds, and thus avoid water and nutrient competition, and to incorporate fertilizers and manure into the soil (Palese et al., 2014a, 2014b). However, conservation tillage showed no negative effects on crop yields because the increase in SOM can positively influence water storage capacity and the availability of nutrients in soil (Almagro et al., 2016; Garcia-Franco et al., 2015). However, after assessing all effects on crop yields with regard to no-tillage or minimum tillage, we have found that, although no overall negative effects have been detected under minimum tillage in the various studies, under arid and semiarid conditions, no-tillage may have negative effects on crop yields in rainfed orchards, especially in almond and olive groves (De Leijster et al., 2019; Hernández et al., 2005; Martínez-Mena et al., 2013; Vignozzi et al., 2019). This negative impact can be related to the high soil compaction and sealing due to low SOM content, leading to a lack of sufficient soil aeration and reduced N mineralization rate (López-Garrido et al., 2014; Martínez-Mena et al., 2013). This soil compaction may induce N limitations in the crop, hamper root development and hinder gas exchange in the roots. Under these conditions, minimum tillage is recommended to break soil compactness, unless until SOM is increased so that no compaction is produced (López-Garrido et al., 2014; Martínez-Mena et al., 2013).

Crop yield differences between organic and conventional farming have been reported to average 20% (Lee et al., 2019), a difference that is normally explained by the difficulty of managing soil P in organic systems (Oehl et al., 2002). The lack of a significant negative effect of organic fertilization compared to inorganic fertilization on crop yield might therefore be related to the absence of a negative influence of organic fertilization on available P in soil (Fig. 2). Soil available P content was significantly correlated with crop yields using all the experiments included in this meta-analysis (R = 0.34; P < 0.01), highlighting the importance of soil P on land productivity. Thus, under current management practices in Mediterranean orchards, nutrients such as N and P are not reduced with the adoption of organic fertilizers, while also maintaining the same yields. Therefore, increasing soil organic matter leads to no economic losses expected for farmers in addition to positive effects on soil quality.

Texture is one of the most important variables of soil because it can affect gas exchange, thermic conductivity, aggregates and root growth. Moreover, clay enhances water retention in soils and has a high capacity for cation exchange (Nguyen and Marschner, 2014). Therefore, soils with low clay content tend to be associated with low availability of nutrients and water for plant growth (Zornoza et al., 2015). Our study has detected that soil clay content can influence the response of crop yield to intercropping, conservation tillage and organic fertilization since soils with lower clay content were related to higher response ratios in crop yield than soils with high clay content (Fig. 4). Thus, in soils with low clay content and the associated lower potential fertility, the effects of intercropping, conservation tillage and organic fertilization are more evident than in soils with higher potential fertility, likely associated with the improvements in soil structure and fertility by the adoption of these practices, as discussed in Section 4.1.

It is important to highlight that the consulted articles only reported data on the tree yield, without specifications about complementary yields with cultivation of secondary crops in the alleys (intercropping systems) nor the effects on the land equivalent ratio (LER). LER is used to characterize land use efficiency by representing the sum of the relative yields in an intercrop compared to the sole crop yields. When LER > 1, the intercropping system is advantageous since it produces more yields with the same land requirements (Mead and Willey, 1980). Thus, for the study of yield from all crops in the orchard, the analysis of LER is needed to conclude that intercropping has actual detrimental effects on crop yield since the overall productivity of the land could be enhanced. In this sense, Bai et al. (2016) reported LERs of 1.34, 1.44 and 1.33 in mixed systems between rainfed apricots and peanuts, millet and sweet potato, respectively, in a semiarid area of China. Arenas-Corraliza et al. (2018) also reported an LER > 1.34 in a hybrid walnut silvoarable system intercropped with wheat and barley in Spain. These authors concluded that intercropping systems may improve the productivity of rainfed agriculture more than sole pure plantations, especially when a drought adapted crop is used. However, it is true that many cover crops are just grown for environmental benefits, so the LER concept should be adjusted for the value of the product in orchards to make sense as a management decision-making criterion.

Social services were not assessed in this meta-analysis. However, the adoption of diversified cropping systems in Mediterranean orchards, with high diversity and aesthetic value, can attract tourism to rural areas and provide an opportunity for farmers to diversify their incomes (Brandth and Haugen, 2011). For example, almond and olive groves intercropped with flowers such as legumes or aromatics in the alleys (Belmonte et al., 2018; Hugo et al., 2008; Novara et al., 2019; Rodrigues et al., 2015) can increase the aesthetic value of the landscape, attracting tourism and enhancing the profitability of the agroecosystems, as in the case of the High Provence in France, where lavender has become a tourist attraction (Lorenzini, 2010).