Conducting an in situ evaluation of erodibility in a Mediterranean semi‑arid and conventional vineyard in Granada province (Southern Spain) through rainfall simulation experiments

Vol.:(0123456789) Euro-Mediterranean Journal for Environmental Integration (2024) 9:797–808 https://doi.org/10.1007/s41207-024-00485-4 CONFERENCE PAPER Conducting aninsitu evaluation oferodibility inaMediterranean semi‑arid andconventional vineyard inGranada province (Southern Spain) throughrainfall simulation experiments JoséParraOrtega1· LauraCambronero1· MaríaAlcarriaSalas1· JoséLuisRodríguez1· VíctorHugoDurán‑Zuazo2· SaskiaD.Keesstra1,3,4· JesúsRodrigo‑Comino1 Received: 30 October 2023 / Accepted: 4 February 2024 / Published online: 16 April 2024 © The Author(s) 2024 Abstract Vineyards in Europe has been fundamental for food, drink and cosmetic production, and job creation; however, in recent decades due to increased cultivation intensity, numerous negative consequences, including erosion, have been observed. Erodibility or susceptibility to erosion is a little-studied parameter in vineyards but is crucial for analyzing the vulnerability of this crop. Therefore, in this research, a small portable rainfall simulator was used as a useful tool for assessing erodibility, combined with other methods such as soil analyses or infiltration measurements in a semi-arid vineyard located in the Granada province (Spain) considering 20 different hotspots at diverse hillslope positions along the inter-rows and close to the traffic roads. The experiments were conducted in spring 2022 under dry soil conditions. Our results display susceptibility to erosion, particularly on steeper parts such as the shoulder and backslopes. In these areas, runoff gained momentum, carrying a significant sediment load, diminishing the effectiveness of stone cover, and occasionally leading to its removal, especially near the roads. Nevertheless, it is observed that increased roughness plays a mitigating role by slowing down runoff. Using linear correlation analysis and Spearman rank coefficient, we observed this effect is linked to factors such as stoniness, vegetation, and moderate tillage. Conversely, in the flatter zones, primarily in lower areas, reduced runoff and delayed onset are primarily influenced by factors such as roughness, type of cover, material composition, and organic matter content. We concluded that this study case can demonstrate that erodibility in vineyards can shed light as an extra parameter to inform farmers, rural inhabitants, and policymakers about the extreme problem of the vulnerable soils of vineyards. Keywords Soil erodibility· Soil erosion· Viticulture· In situ experiments· Hydrological processes Responsible Editor: Riheb Hadji. * Jesús Rodrigo-Comino [email protected] Laura Cambronero [email protected]r.es María Alcarria Salas [email protected] José Luis Rodríguez [email protected]r.es Víctor Hugo Durán-Zuazo [email protected] Saskia D. Keesstra saskia.keess[email protected] 1 Departamento de Análisis Geográfico Regional y Geografía Física, Facultad de Filosofía y Letras, Campus Universitario de Cartuja, University ofGranada, 18071Granada, Spain 2 IFAPA Centro “Camino de Purchil”, Camino de Purchil s/n, 18004Granada, Spain 3 Climate-KIC Holding B.V. Plantage, Middenlaan 45, Amsterdam, TheNetherlands 4 Team Soil Water andLand Use, Wageningen Environmental Research, P.O. Box47, 6700AAWageningen, TheNetherlands 798 Euro-Mediterranean Journal for Environmental Integration (2024) 9:797–808 Introduction Agriculture has been an essential economic sector in Europe for centuries. It has played a crucial role in providing food for the population as well as creating many jobs. Over the past decades, especially during the 21st century, it has taken on a more commercial and technological focus. For this reason, most of the land is dedicated to agriculture; the area of cultivated and plowed land accounts for approximately25% of Europe considering the last reports of the European Environmental Agency. Over the years, there has been an attempt to maximize production, which has led to an intensification of cultivation; however, this has been affected by inadequate soil management, resulting in serious environmental consequences such as increased erosion and runoff (Boardman etal. 2003; Cerdà etal. 2009; Kairis etal. 2013). Soil erosion is an environmental issue that affects many countries, including Spain, and can have serious consequences both for agriculture and the local economy (García-Ruiz 2010). Viticulture is an agricultural activity, especially important in a country like Spain. It is the world's third-largest wine producer, behind countries such as Italy and France, although production is lower compared with previous years (International Organisation of Vine and WineIntergovernmental Organisation 2019). Despite being third in production, Spain ranks at the top in terms of the largest vineyard cultivation area globally. In Andalusia, due to its lithology, topography, and climate, vineyards are found on steep slopes, old embankments, or between different levels of embankments and river terraces. Additionally, the soils in these areas are often bare without vegetative cover for much of the year, characterized by a high percentage of rock fragments (Rodrigo-Comino etal. 2017). It is worth noting that these crops often undergo excessive tillage, causing the soils to become bare and exposed to rainfall, leading to raindrops further eroding the soil and triggering surface runoff, displacing the eroded material (Martínez-Casasnovas etal. 2005; Parras-Alcántara etal. 2016; Taguas etal. 2013). Soil erodibility is a parameter that has been under studied in Spanish vineyards, especially in Andalusia, which is defined as a measure of soil susceptibility to erosion and is influenced by factors such as texture, structure, porosity, and chemical composition (Auerswald etal. 2014; Bryan 2000). Soils with low erodibility are less susceptible to erosion, while soils with high erodibility are more prone to soil and nutrient loss (Choi etal. 2016; Nemetova and Danacova 2019; Parlak etal. 2016). The relationship between erosion in Spanish vineyards and soil erodibility, if further analyzed, would be key to understanding and addressing potential solutions to mitigate it. However, the scientific and technical literature does not provide clear insights into this, at least not with insitu experiments, especially in vineyards. There is a recent publication at the European scale trying to shed light on it but the resolution is large and is not divided into land uses (Panagos etal. 2014). In situ measurement of soil erosion is a complex issue, and various techniques exist for its evaluation, but they require long monitoring periods. One of the most used methods is the direct measurement of the amount of soil eroded in a specific area, using techniques such as measuring changes in soil profile height before and after erosion (Rodrigo-Comino etal. 2019) or measuring the amount of sediment and water deposited in boxes or containers placed at the base of the plot along the main slope direction (Biddoccu etal. 2017). In terms of erosion measurement and erodibility modeling, it can also be predicted using mathematical models such as the Water Erosion Prediction Project (WEPP) (Flanagan etal. 2001) or the Universal Soil Loss Equation (USLE) (Wischmeier and Smith 1965) and its revision (RUSLE) (Renard etal. 1991). Another option is to measure erosion in the field using runoff or rainfall simulators, which allow for field experiments to quantify erosion and its activation process (Cerdà 1999). The portable rainfall simulator is a device used to simulate artificial rainfall in a specific area with controlled conditions such as intensity, duration, kinetic energy, drop size, etc. (Bryan and De Ploey 1983; Cerdà 1998a). Its main purpose is to measure the process of soil erosion activation in the field or under laboratory conditions. Simulators can be of various types, which, according to (Iserloh etal. 2013), are classified on the basis of the nozzle type, such as those developed in Tübingen, Córdoba, or Basel; on the basis of capillary type, developed in Granada and Wageningen; and on the basis of circular plot diffuser type with a base, developed in Almería, Málaga, Murcia, Trier, Zaragoza, Valencia, Zaragoza University, and La Rioja. Mini-rainfall simulators are also a type of portable simulator with a very small analysis area, and the Eijkelkamp model from the Netherlands stands out. This simulator is a very useful tool as it allows for numerous experiments to be conducted in less time, with less water consumption and fewer personnel, reducing costs and effort. However, its representativeness is lower due to its study area and only allows for the estimation of soil erodibility (Danáčová etal. 2017). Nevertheless, for this study, this property is the ultimate research goal. This type of simulator would be the first time it is used to measure erosion and erodibility in Spanish and Andalusian vineyards, making it an innovative method that could provide complementary results for future research and help address the issue of soil erosion and its susceptibility. To complete these types of studies, it is also common to use other instruments such as infiltrometers, which allow for 799Euro-Mediterranean Journal for Environmental Integration (2024) 9:797–808 measurements of water velocity through the surface horizons of the soil (Naik and Pekkat 2022). Infiltrometers consist of a cylinder subjected to a water load, and the volume of liquid drained per unit of time is measured, known as infiltration capacity. There are different types of infiltrometers, including cylinder or inundation infiltrometers (single or double), closed infiltrometers, or tension infiltrometers (Alagna etal. 2016; Di Prima 2015). Therefore, the main goal of this research is to estimate soil erodibility using a mini-rainfall simulator in a vineyard located in the province of Granada. To achieve this, 20 rainfall simulations and 20 infiltration measurements supported by the analysis of key soil properties were conducted. The specific objectives were as follows: (i) estimate the time the soil is capable of retaining water before mobilizing it as surface runoff; (ii) identify factors related to topography or soil properties that affect the activation of runoff and erosion processes; (iii) propose measures to reduce soil vulnerability to erosion to prevent it from affecting the quality and quantity of grape production; and (iv) lay the foundation for a potential new research topic with data that complements the analysis of environmental degradation and the impact of climate change on Mediterranean vineyards. Materials andmethods Study area A total of 20 rainfall simulations and twenty infiltration measurements, supported by the analysis of key soil properties, were conducted between March and April 2023, in a vineyard located in the municipality of Villamena, which belongs to the Valle de Lecrín region, in the province of Granada. The study area is located within the Alpujárride Complex, corresponding to the most recent materials of the mountain chains of Albuñuelas, Almijara, and Guájares, along with the southwestern part of Sierra Nevada. It consists of mica schists, limestone, and dolomite, which are not continuous. The study area is located within the Alpujárride Complex, specifically, it corresponds to the most recent materials of the mountain chains of Albuñuelas, Almijara, and Guájares, along with the southwestern branch of Sierra Nevada. The Alpujárride Beds are classified into two groups, those located to the north of the Pinos del Valle parallel and those to the south; the plot is located in the northern area of Pinos del Valle, between Dúrcal and Jayena. Alcázar Bed, formed by variegated phyllites (violet, gray, green, and reddish colors). The presence of limestones and dolomites is not continuous and is poorly developed. The Cástaras Bed is mainly composed of phyllites alternating with quartzites at the base. The roof exhibits a well-represented carbonate formation. It has a Mediterranean climate (Padul Meteorological Station—Institute of Agricultural and Fisheries Research and Training), affected by high temperatures and severe drought since 2022. From early February to late May, the average temperature is 13.9°C, and rainfall has not exceeded 5mm during these study months (meteorological station installed, which records data from January 2023). The average relative humidity has been 47.1%. Overall, rainfall intensity is moderate, with an absolute maximum in winter, with December being the rainiest month. The second peak in rainfall occurs in spring with little difference from the autumn months. On the other hand, the driest months correspond to summer, with August being the least rainy month. Rainfall simulation experiments For the rainfall simulation experiments in the vineyard, a total of 20 different points were selected (Fig.1) during 4days of fieldwork involving at least three people. Each day, a total of five experiments were conducted to assess the effects of rain on the soil under known and controlled conditions (intensity, drop size, amount of water applied over an area, and duration of the experiment). In this research, the Eijkelkamp model from the Netherlands was used, which consists of a box with a sprinkler system that simulates rainfall intensity and type. The water sprayed by the sprinklers falls onto a specific area of soil, allowing for the measurement of the amount of sediment lost, runoff, and water absorbed by the soil. This simulator is a very useful tool as it allows for numerous experiments to be conducted in less time, with less water consumption and fewer personnel, reducing costs and effort. However, its representativeness is lower due to its study area being less than 0.0625 m2, and in this research, it will allow us to estimated erodibility (Iserloh etal. 2013). The experiment is carried out by discharging an identical volume of water at each point, collecting the resulting water as runoff, and soil loss that flows through the drainage point or outlet. The drop size is 5.9mm, falling during 360s with an intensity of 6mm min−1. The miniportable rainfall simulator consists of three main parts and some accessories (Fig.2a, b). The first part consists of the showerhead, which has an internal regulator that generates uniform rainfall through 49 tubes working as capillary pressure flow. The second part is the adjustable aluminum stand on which the showerhead is placed, allowing to use a specific height (in this case, 35–40cm). Finally, there is a frame to place the adjustable stand on top of it, ensuring its stability and preventing water from escaping from the sides, thus aiding in the flow toward the outlet or channel. To conduct rainfall simulation experiments, it is recommended to carry out them under dry conditions to avoid extreme differences in previous soil moisture content as demonstrated by other authors (Cerdà 1998a, b). 800 Euro-Mediterranean Journal for Environmental Integration (2024) 9:797–808 Before starting each experiment, some data were recorded to assess which factors are influencing soil erodibility and runoff activation (Fig.2b): (i) vegetation cover (%) estimated using a grid cell with vegetation visible in a photograph (Cerdà etal. 1997); (ii) stone cover (%), similar to the previous factor; (iii) surface roughness using the chain method horizontally and vertically; and (iv) slope, measured using a digital clinometer. During each experiment, which lasts approximately 6min, the same person (to avoid measurement bias) should record the following times in seconds at the beginning and at each 1-min interval: (i) time to ponding (Tp) as the time it takes for the entire plot to become wet from the start of water discharge; ii) time to outlet (To) as the exact moment when the first drop flows through the plot; and,(iii) time to runoff generation (Tr) or the exact time at which the first drop reaches the outlet and runoff begins. When the experiment finishes, runoff volume (l; total volume of water collected at the end of the outlet at each interval) and sediment yield (g; material detached by drop impact -splash- and transported by runoff) are determined later in the laboratory. The runoff water should be evaporated at 110°C, and the total solids are then weighted. Finally, sediment concentration can be obtained dividing sediment yield by runoff (g l−1). Infiltration measurements At the same time that rainfall simulations are carried out, infiltration measurements were taken using a mini-disc infiltrometer. A total of 20 repetitions were performed in Fig. 1 Localization of the study area and experimental plots Vineyard (Bodegas Calvente) Rainfall simulation experiments and infiltration experiments. Source: PNOA (IGN, 2020) Fig. 2 Field campaign using the mini-rainfall simulator (a, b) and mini-disc infiltrometer(c)ABC 801Euro-Mediterranean Journal for Environmental Integration (2024) 9:797–808 near areas where the rainfall simulations took place. The experiment results provide data on hydraulic conductivity and infiltration rates. In Fig.2c, the mini-disc infiltrometer, which consists of an acrylic tube with a semi-permeable disk at the bottom and a tube at the top for regulating suction can be observed. The plastic tank is filled, setting the suction (which in this case was positioned at 4.5cm), and measurements are taken at regular intervals (10s, 30s, 1min, 2min, and 5min) with the help of a stopwatch. Related key soil properties Once each rainfall simulation was completed, soil samples (between 0.5 and 1kg; n = 20) were collected from three points near the experiment. All samples were numbered and stored in airtight bags for transportation to the laboratory. In the laboratory, they were sieved, separating the fine material (< 2mm) from the coarse one (> 2mm). Subsequently, organic matter analysis was carried out in a muffle furnace at a temperature of 430°C, following the ignition estimation method (Ball 1964; Rather 1918). Furthermore, pH was calculated by direct measurement using a pH meter. To do this, samples must be prepared by diluting them in distilled water. Once calibrated, the pH of each sample was measured by inserting the electrode into the diluted sample. Statistical analysis The data obtained were organized and normalized. Then, the data were encoded in Excel data sheets (Microsoft, USA). Firstly, average values, standard deviations, and maximum/ minimum values were estimated. The results were analyzed using scatter plots to observe linear regressions. Subsequently, using SPSS software (IBM, USA), correlations were conducted using the Spearman method to observe if there were statistically significant relationships (p < 0.001 and p < 0.5) after the normalization of the dataset. This coefficient was used when non-linear relationships are expected instead of other coefficients such as the Pearson correlation. Results Soil properties andplot characteristics In Fig.3, the key soil properties of the plots are depicted. The experimental area are characterized by more than 54% of gravels (> 2mm). The overall averages for organic matter and pH are 6.39 and 7.5, respectively. The organic matter values range from 3.86 to 11.13, corresponding to samples 13 (zone 3, road) and 14 (zone 3, inter-row). The pH values Fig. 3 Soil properties and environmental characteristics of the rainfall simulation plots 802 Euro-Mediterranean Journal for Environmental Integration (2024) 9:797–808 vary between 6.8 and 8.0, corresponding with samples 6 (zone 2, road) and 7 (zone 2, inter-row). The average percentage of vegetation cover and rock fragments were 1.6% and 10.7%, respectively. The maximum value for vegetation cover is 4.7%, and for rock fragments at the surface, it is 33.9%. Regarding the minimum values, rock fragments have a minimum of 2.9%, while vegetation was 0%, especially in the roads. As for the slope, the average value was 3.1°, with a maximum of 7° and a minimum of 0.5°. The maximum slope was observed in the upper zone of the vineyard, while the minimum was in the backslope zone. Finally, roughness averaged 0.9mm mm−1 with minimum values reaching 0.8mm mm−1. Infiltration measurements The measurements conducted with the mini-disc infiltrometer are showed in Fig.4. Specifically, in Fig.4a, the results are depicted per intervals and in Fig.4b per measurement on the field. During the first minute, the average infiltration rates per interval overpassed 7mm s−1, then, it decreased to less than 4mm s−1. From the second minute, with wider intervals (30s), the average rates are 2.5mm s−1, and then it decreases from the tenth minute to 1.1mm s−1. Then, the steady state is reached averaging 0.4mm s−1. The highest infiltration measurements were the sample points 5, 13, and 15, with more than 3mm s−1, all of them situated in the shoulder and backslope parts within the inter-row areas. On the contrary, the lowest average infiltration rates were recorded at sampling points 1, 10, and 11, corresponding with the same hillslope positions but situated along the roads. Rainfall simulation results In Fig.5, box plots shows the hydrological responses during the initial time of all rainfall simulation experiments. The average time for Tp was 33.7s, ranging from a maximum of 65s to a minimum of 20s. The duration until the first drop reaches the drain (To) averaged at 62.1s, with a maximum of 162s and a minimum of 35s. Finally, the time it takes for runoff to generate (Tr) recorded an average value of 165.1s, ranging from a maximum of 320s to a minimum of 0s, reflecting a case where all the water infiltrated, and no runoff occurred. In Figs.6 and 7 scatter plots depict the results of total volume and soil loss in relation to environmental plot characteristics and infiltration measurements. Concerning runoff, the average volume was 0.13l, with a maximum of 1.17l (sampling point number 3, on the road) and a minimum of 0 (sampling point number 2, near a vine in the inter-row area), where no runoff occurred in one rainfall simulation. Regarding sediment contribution, the average value was 0.90g of solids, with a maximum of 9g and a minimum of 0g, similar to the patterns observed in the same sampling points (3 and 2). These results corresponded to the highest average sediment concentrations in sampling points 3 Fig. 4 Average infiltration rates per Interval and total values 0 1 2 3 4 5 6 7 8 9 smm(etarnoitartlifnI -1) Time 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 1234567891011121314151617181920 mm s-1 n a b Fig. 5 Hydrological response during the rainfall simulation experiments 803Euro-Mediterranean Journal for Environmental Integration (2024) 9:797–808 (11.3g l−1) and 11 (22.3g l−1), both situated along the roads. In the inter-row areas, the highest average sediment concentrations ranged from 8.7 to 8.8g l−1 in sampling points 15 and 9, respectively. Finally, scatter plot graphs register that an increase in slope (degrees) correlates with an increase in runoff and sediment contribution. There is also an impact of increased stone coverage on overall values in runoff and soil loss. In supplementary material 1, a table summarizing all the results of each simulation can be found. Correlation amongsoil erodibility results andenvironmental plot characteristics In Table1 and 2, multiple correlation analyses, specifically Spearman rank coefficients, were conducted. Initially, total soil erodibility results were considered, showing the highest significant correlations between roughness and stone cover with time to outlet, reaching –0.52 and –0.53, respectively. In the inter-row areas of the rainfall simulation experiments, the highest correlations were observed between < 2mm or fine soil grain size fraction and Tp (–0.51) and R (0.50). For experiments conducted on roads, the absence of vegetation cover led to a decrease in To (0.61) and Tr (0.74), accompanied by an increase in sediment yield (–0.67) and runoff (–0.79). Additionally, roughness showed a significant correlation with To (–0.61) and Tr (–0.57). Moreover, as the slope increased, Tp recorded a reduction (–0.65). Focusing on correlations per sampling point area, in the shoulder, an increase in stone cover generated a positive correlation with Sy (0.90), R (0.90), and Sc (0.90). Similar R² = 0.0841 0 0.2 0.4 0.6 0.8 1 1.2 1.4 0 1 2 3 4 5 Runoff (l) Vegetation cover (%) b R² = 0.1808 0 0.2 0.4 0.6 0.8 1 1.2 1.4 02468 )l(ffonnuR Slope (º) a R² = 0.4694 R² = 0.1086 0 0.2 0.4 0.6 0.8 1 1.2 1.4 02040608 01 00 Runoff (l) Stone cover (%) Rock fragment cover (%) >2mm <2mm R² = 0.0976 0 0.2 0.4 0.6 0.8 1 1.2 1.4 0.80.820.840.860.880.9 0.92 Runnoff (l) Roughness (cm cm -1 ) e R² = 0.0108 0 0.2 0.4 0.6 0.8 1 1.2 1.4 024681012 )l(ffonnuR Organic matter (%) d R² = 0.0047 0 0.2 0.4 0.6 0.8 1 1.2 1.4 01234 Runnoff (l) Infiltration (mm) f c Fig. 6 Scatter plots depicting total runoff volume and conditioning factors related to the plot characteristics R² = 0.0588 0 2 4 6 8 10 012345 Soil loss (g) Vegetation cover (%) b R² = 0.0354 0 2 4 6 8 10 0.80.820.840.860.880.9 0.92 Soil loss (g) Roughness (cm cm-1) e R² = 0.0816 0 2 4 6 8 10 02468 )g(ssollioS Slope (º) a R² = 0.0345 0 2 4 6 8 10 0510 )g(ssollioS Organic matter (%) d R² = 0.0018 0 2 4 6 8 10 01234 Soil loss (g) Infiltration (mm) f R² = 0.3023 R² = 0.0386 0 2 4 6 8 10 02040608 01 00 Soil loss (l) Stone cover (%) Rock fragment cover (%) >2mm <2mm c Fig. 7 Scatter plots depicting total soil loss and conditioning factors related to the plot characteristics 804 Euro-Mediterranean Journal for Environmental Integration (2024) 9:797–808 results were obtained with an increase in Rg and soil erodibility (0.87). Furthermore, when roughness (–0.90) and slope (–0.95) decreased, To increased. In the upper-back- slope part, < 2mm influenced Tp (–0.90) and Sy (0.90), which also correlated with Rg (–0.98). In the rainfall experiments performed in the low-backslope part, an increase in Sl obtained a significant correlation with an increase in Tp (0.98), and the absence of Vc implied higher Sy and Sc. Finally, in the footslope, the presence of vegetation cover or high contents of organic matter did not reduce runoff and Tr. Only a higher content of stones at the surface and along the first 20cm showed any influence on Sy and R. Discussion The theme addressed in this research is under explored in Spanish vineyards and, to a lesser extent, in Andalusia. Hence, we have attempted to go further into the relationship between erosion in vineyards and soil erodibility. Understanding and addressing this understudied parameter would be crucial in mitigating erosion and finding effective solutions (Al-Hamdan etal. 2017; Ayoubi etal. 2018). In the experiments conducted in Bodegas Calvente’s vineyard plot, rainfall simulations showed relatively homogeneous values throughout the parcel, but also it was indicated specific vulnerable points in the vineyard. Factors influencing vulnerability include stone cover and soil roughness, both of which slow down the runoff generation process (Seeger 2007). An increase in slope also results in higher runoff volumes. According to traditional studies conducted under laboratory conditions (Aksoy and Kavvas 2005; Bryan 2000; De Ploey 1991), erosion is more pronounced on steeper slopes due to increased water velocity and particle movement. This effect is even more pronounced on roads (Salesa etal. 2019). While slope has been studied for decades as a trigger for erosion, the influence of roads, especially in vineyards, remains to be thoroughly evaluated, as also mentioned by Rodrigo-Comino etal. (2015) in German vineyards close to steep vineyards and the taluses. Regarding stoniness, Poesen etal., (1998) suggest that rock fragments increase the time of runoff concentration and decrease its volume compared to bare soil surfaces. A stony surface may promote faster infiltration and deeper penetration of applied water due to the contact between stones and soil matrix, facilitating quicker and deeper flow (Jomaa etal. 2013, 2012). However, rock fragments may reduce infiltration rates and increase runoff generation if embedded in the upper layer. However, in some parts of the hillslope, rock fragments even increased soil erodibility, which totally Table 1 Spearman rank coefficient between soil erodibility results and environmental plot characteristics(total and considering inter-row or close to the roads) *p > 0.5; **p > 0.01 Vc vegetation cover, Rg roughness, Scover stone cover, Sl slope, > 2mm gravels, < 2mm fine material, OM organic matter, Tp time to ponding, To time to outlet, Tr time to runoff, Sy sediment yield, R runoff, Sc sediment concentration Total TpToTr Sy R Sc Total Vc 0.19 0.09 –0.20 –0.10 –0.09 –0.14 Rg –0.02 –0.52*–0.29 0.28 0.34 0.22 Scover –0.09 –0.53*–0.05 0.09 0.30 0.02 Sl –0.02 –0.02 –0.26 0.12 0.35 0.04 > 2mm –0.20 –0.07 0.17 0.10 0.15 0.19 < 2mm –0.28 0.09 0.20 0.08 0.09 0.20 OM 0.14 0.22 –0.35 –0.07 0.03 –0.17 Inter-row areas Roads Vc 0.20 0.43 –0.37 –0.26 – 0.23 – 0.31 Rg – 0.05 – 0.45 – 0.05 0.29 0.25 0.19 Scover –0.19 –0.25 0.49 –0.01 –0.02 –0.06 Sl 0.02 0.23 –0.13 0.08 0.29 0.08 > 2mm –0.44 0.07 0.40 0.12 0.26 0.35 < 2mm –0.51 0.01 0.14 0.24 0.50 0.39 OM 0.34 0.20 –0.43 –0.07 0.04 –0.24 Vc –0.35 0.61 0.74 –0.67 –0.79*0.36 Rg 0.13 –0.61 –0.57 0.32 0.46 –0.07 Scover –0.33 –0.40 –0.16 0.49 0.58 0.11 Sl –0.65 0.00 –0.04 –0.25 –0.36 –0.54 > 2mm 0.11 0.18 0.18 –0.11 –0.32 0.00 < 2mm –0.35 0.46 –0.21 0.07 0.18 0.21 805Euro-Mediterranean Journal for Environmental Integration (2024) 9:797–808 contradict the results of other investigations conducted in Mediterranean vineyards characterized by similar environmental conditions and using small portable rainfall simulations (Rodrigo-Comino etal. 2017). In the results obtained by vineyard hillslope parts, various parameters stand out. In the shoulder, the highest and steepest part of the vineyard, both roughness and stone cover play significant roles. Higher roughness delays runoff, aligning with general results. With increased slope, the time for runoff initiation decreases. Stone cover enhances concentration, especially when coupled with high roughness, which correlates with the slope. In the rest of the parts, roughness, vegetation cover, and stone cover are closely related. In these flatter areas, both vegetation and stone cover reduce runoff, delaying its onset, and decreasing sediment contribution. This happens because, in flat vineyard areas, runoff moves more slowly, and higher roughness and greater cover percentage slow down its surface movement (Bagagiolo etal. 2018). These factors can be linked to erodibility, i.e., the susceptibility to erosion (Wang etal. 2019). To reduce erodibility in the vineyard, increasing soil roughness and coverage with either vegetation or stones would be essential to minimize runoff and, consequently, erosion. The challenge with vegetation cover is potential competition with the grapevines for nutrients and water (Marques etal. 2021; Ruiz-Colmenero etal. 2011), although water may be less Table 2 Spearman rank coefficient between soil erodibility results and environmental plot characteristics considering sampling plot areas(hillslope positions) * p > 0.5; **p > 0.01 Vc vegetation cover, Rg roughness, Scover stone cover, Sl Slope, > 2mm gravels, < 2mm fine material, OM organic matter, Tp time to ponding, To time to outlet, Tr time to runoff, Sy sediment yield, R Runoff, Sc sedimend concentration Sampling points TpToTr Sy R Sc 1–5 Vc –0.39 0.03 0.56 0.15 0.15 0.15 Rg 0.29 –0.90*–0.21 0.87 0.87 0.87 Scover –0.36 –0.87 0.50 0.90*0.90*0.90* Sl –0.13 –0.95*0.21 0.87 0.87 0.87 > 2mm –0.56 –0.15 0.70 0.20 0.20 0.20 < 2mm –0.41 –0.15 0.50 0.10 0.10 0.10 OM 0.72 0.10 –0.80 –0.30 –0.30 –0.30 6–10 Vc 0.20 –0.50 –0.10 –0.20 –0.22 0.10 Rg 0.72 –0.10 0.41 –0.98** –0.34 –0.82 Scover 0.50 –0.60 –0.10 –0.70 0.22 –0.60 Sl 0.20 0.10 –0.10 –0.20 0.45 –0.50 > 2mm –0.50 0.30 –0.20 0.60 0.45 0.30 < 2mm –0.90*0.00 –0.30 0.90*0.45 0.80 OM 0.70 0.40 0.10 –0.30 –0.45 –0.40 11–15 Vc 0.34 0.67 0.45 –0.89*–0.46 –0.89* Rg –0.20 –0.70 –0.30 0.30 0.21 0.30 Scover 0.70 0.20 0.30 –0.30 –0.21 –0.30 Sl 0.98** 0.72 0.87 –0.62 –0.76 –0.62 > 2mm –0.20 0.20 0.30 0.00 –0.36 0.00 < 2mm –0.70 –0.20 –0.70 0.20 0.67 0.20 OM 0.30 0.00 –0.30 0.20 0.46 0.20 16–20 Vc 0.18 –0.18 –0.71 0.35 0.71 0.00 Rg –0.53 –0.53 –0.21 0.36 0.21 0.67 Scover 0.56 0.36 0.90*–0.80 –0.90*–0.50 Sl –0.03 0.39 0.05 0.21 –0.05 0.56 > 2mm –0.36 0.36 –0.80 0.90*0.80 0.80 < 2mm 0.15 0.05 –0.80 0.60 0.80 0.50 OM –0.36 –0.98** –0.10 0.00 0.10 0.10