The impact of Macrochloa tenacissima and Anthyllis cytisoides on soil structural stability - A field study in the Rambla Honda Basin, Almería Province, Spain

(1)

The impact of Macrochloa tenacissima and Anthyllis

cytisoides on soil structural stability

A field study in the Rambla Honda Basin,

Almería Province, Spain

Figure 1. (De Boer, 2018)

Niels Verweij – 10690514

BSc Thesis Project

Bèta-Gamma – major Earth Sciences

University of Amsterdam – Institute for Biodiversity and Ecosystem Dynamics (IBED)

Supervisor: Dhr. dr. L.H. (Erik) Cammeraat

(2)

2

Abstract

Soil degradation as a result of agricultural land abandonment is one of the bigger environmental problems in southeastern Spain. This causes a loss of valuable agricultural fields and a higher risk of flash floods in the surrounding villages and cities. It is necessary to find measures to counter this ongoing process. A suggested measure is active replanting of Anthyllis cytisoides and

Macrochloa tenacissima as this improves soil properties and reduces soil erosion. This research

assesses the influence of Anthyllis cytisoides and Macrochloa tenacissima on soil structural stability in the Rambla Honda Basin in southeastern Spain. Aggregate stability, soil organic matter, inorganic and organic carbon content were measured in samples from different locations with respect to the catena and vegetation patches. Analysis was done in field and in laboratory environment and included the Herrick field test, CND analysis, CNS analysis and the Van Wesemael method. Results show that when pooled together Anthyllis cytisoides and Macrochloa

tenacissima have a positive influence on soil organic matter and aggregate stability. The influence

of Macrochloa tenacissima differs for the two locations along the catena, having a greater impact on SOM and aggregate stability on the hillslope than on the alluvial fan. The influence of Anthyllis

cytisoides does not differ significantly between the hillslope and the alluvial fan. The total carbon

shows overall similar results to soil organic matter. Inorganic carbon is low for all samples and shows no significant differences between any sample locations. Anthyllis cytisoides and

Macrochloa tenacissima have a positive influence on soil structural stability.

This research suggests that after abandonment of agricultural fields replanting with Anthyllis

cytisoides on the alluvial fan and Macrochloa tenacissima on the hillslope will reduce soil erosion

risk, by increasing soil organic matter and aggregate stability.

Keywords: land abandonment, soil structural stability, soil degradation, erosion, vegetation cover, Anthyllis cytisoides, Macrochloa tenacissima, Spain, Rambla Honda Basin

(3)

3

1. Introduction

In semi-arid regions like southeastern Spain abandonment of agricultural fields and overgrazing are widely found phenomena (Cerdà, 1997, 1998). In a natural environment an ecosystem is able to restore itself after such events due to the ecosystem’s resilience. However, when human interference plays a role in the system, the resilience of an ecosystem could not be strong enough to bring the system back in balance. This means the system can collapse and cause degradation of soil, a decline in biomass production and erosion of soil.

In southeastern Spain large scale agricultural land abandonment occurred during the 1950s and 1960s (Cerdà, 1997). This caused large plots of land to become unregulated and unirrigated. As a result of this, soil degradation and soil erosion started to occur. The first years after the vegetation removal in an ecosystem the organic carbon content (OC) in the soil stay level, but later OC and the percentage of stable aggregates dropped significantly (Albaladejo et al., 1998; Martínez-Mena

et al., 2002). This process causes one of the bigger environmental problems in Spain.

Aggregate stability is generally used as an indicator of soil structural stability (Cammeraat, 2004; Bronick and Lal, 2005). It is measured relatively easily and to a low cost. Soil aggregates are clusters of soil particles and bound together more strongly than to other particles. The stability of these aggregates is one of the factors that determines soil stability and is a key-indicator for understanding the resilience of an ecosystem to disturbance (Cammeraat and Imeson, 1998; Cantón et al., 2009). Changes in aggregate stability are an early indicator of soil degradation or recovery (Cammeraat and Imeson, 1998). Clustering of soil aggregates is caused by many soil properties, for example the type and texture of weathered plant material in the soil (Cammeraat and Imeson, 1998). Biota, clay and carbonates also influence the aggregate stability (Bronick and Lal, 2005).

When looking at the differences in soil properties along a slope it is found that soil moisture content (SMC), soil organic matter (SOM) and vegetation cover are higher at footslope positions than at backslope positions on eroded rangeland in semi-arid environments (Oztas, Koc and Comakli, 2003). However without the influx of organic materials from the backslopes, footslopes may become nutrient-depleted because of microbial immobilization (Norton, Sandor and White, 2003). This emphasizes the importance of the connectivity between hillslopes and alluvial soils. Vegetation cover has a positive influence on SOM and soil aggregation (Cammeraat and Imeson, 1998). The difference of impact between Anthyllis cytisoides and Macrochloa tenacissima on soil structural stability has been measured in prior research. Significant results between Macrochloa

tenacissima and Anthyllis cytisoides have been found for runoff and soil loss during rain events

(Bochet, Poesen and Rubio, 2006). However, no significant difference was found for aggregate stability and soil organic material (Cantón et al., 2009). Cammeraat & Imeson (1999) found that the area around the stem of Macrochloa tenacissima has high biological activity and an active root system. The area under the canopy is slightly water repellent. Outside the canopy border a bare crusted area embedded with small rocks is present.

This research investigates the impact of Macrochloa tenacissima and Anthyllis cytisoides on soil structural stability and assesses whether replanting of those vegetation species on abandoned agricultural field reduces the risk on soil erosion.

(5)

5

2. Study area

The study area for this research is the Rambla Honda Basin. It is located near Tabernas in the Almería Province, southeastern Spain (Fig. 2). The mean annual precipitation is 265 mm and mean annual temperature is 17.8 °C (Cantón et al., 2009). Bedrock mainly consists of micha schist with an abundance of quartz veins and phyllite layers and soils mainly consist of sandy loam (Nicolau

et al., 1996; Cantón et al., 2009).

Vegetation grows in patches along the catena, with bare spots between the plants. Macrochloa

tenacissima, which has a shallow rooting system, dominates the vegetation at the top of the

hillslope. Anthyllis cytisoides, which has a deep vertically oriented rooting system, dominates the vegetation at the alluvial fan. (Nicolau et al., 1996; Bochet, Rubio and Poesen, 1999). The slope angle of the plots located on the hillslope and alluvial fan are approximately 20° and 9°, respectively.

Figure 2. Location and overview of study area (adapted from Cantón et al., 2009)

3. Research questions

This research aims to investigate the impact of different vegetation species on soil structural stability in southeastern Spain. This was done by examining aggregate stability and SOM of soil samples from the Rambla Honda catchment area. This research focuses on Anthyllis cytisoides and

Macrochloa tenacissima as these are dominant vegetation species in the research area

(Puigdefábregas et al., 1998). Besides the differences in soil quality between the Anthyllis

cytisoides and Macrochloa tenacissima, this research also examines the differences in soil quality

between soil located on the hillslope and on the alluvial fan. To investigate this the following research question is answered:

What is the impact of Anthyllis cytisoides and Macrochloa tenacissima on soil structural stability in the Rambla Honda Basin?

To answer the main research question the following sub-questions are answered: - What is the impact of SOM on aggregate stability?

- How differs the impact of Anthyllis cytisoides and Macrochloa tenacissima on soil structural stability?

- How differs the impact of Anthyllis cytisoides and Macrochloa tenacissima on soil structural stability between the hillslope and alluvial fan?

(6)

6

4. Hypotheses

Research has shown that soil aggregation is positively influenced by organic matter (Cammeraat and Imeson, 1998; Cerdà, 1998). It is expected that SOM has a positive influence on aggregate stability.

Research has also shown that vegetation cover has a positive impact on SOM and soil aggregation because vegetation patches form a resource island in a more bare landscape (Boix et al., 1995; Cammeraat and Imeson, 1998; Cantón et al., 2009; van Hall et al., 2017). It is expected that SOM and aggregate stability are positively influenced by Macrochloa tenacissima and Anthyllis

cytisoides.

Oztas et al. (2003) found that concentrations of SOM are higher at footslope positions than at upslope positions. However, other research has found that in the Rambla Honda Basin SOM increased upslope and found larger amounts of plant litter under Macrochloa tenacissima than under Anthyllis cytisoides (Nicolau et al., 1996; Puigdefábregas et al., 1996). Macrochloa

tenacissima is dominant on the hillslope, resulting in higher concentrations SOM on the hillslope.

Because of the slope angle the soil on the hillslope has a downslope movement, causing the SOM rich soil to spread out over the hillslope. It is expected to find higher concentrations of SOM and more soil aggregation at vegetation patches on the hillslope.

Prior research has found no significant difference between Anthyllis cytisoides and Macrochloa

tenacissima for soil organic matter and aggregate stability (Bochet, Rubio and Poesen, 1999;

Cantón et al., 2009). It is expected to find a similar impact Anthyllis cytisoides and Macrochloa

tenacissima on soil organic matter and aggregate stability, when analyzing pooled results from the

alluvial fan and the hillslope. However, due to its rooting system Anthyllis cytisoides is more dominant on the alluvial fan. This might result in higher SOM and more stable aggregates near

Anthyllis cytisoides on the alluvial fan. On the hillslope higher SOM and more stable aggregates

(7)

7

5. Methodology

5.1 Fieldwork methods 5.1.1 Sampling

Sampling took place on the Rambla Honda catena. The area was divided in five transects. Every transect included four plants, two Macrochloa tenacissima and two Anthyllis cytisoides plants. One of each plant was located on the hillslope and one of each was located on upper part of the alluvial fan. Near every plant three samples were taken: at the stem, at the canopy border and at a bare patch located approximately 50 cm outside the canopy border oriented upslope from the plant. In some cases large surface rocks prevented sampling at bare patches at the most ideal location. In these situations the distance to the canopy border deviated a few centimeters.

A total of 60 samples were retrieved. Replicates per plant were not taken in this research, as the five transacts act as replicates. For several laboratory analyses, replicates were taken from the samples.

Samples were taken at a depth of 2 – 5 cm below soil surface. The soil was sieved once over a combination of sieves: 4.8 mm, 4 mm and 2 mm. From the 4 – 4.8 mm soil fraction a total of 50 – 60 aggregates were selected for aggregate stability analysis, both in the field and in the laboratory. Testing in the field was done to overcome the risk of damaging the aggregates during transport and thus preventing laboratory analysis. However, this method cannot replace careful laboratory-based measurements (Herrick et al., 2001). Therefore the aggregate stability was also determined in a laboratory. Approximately one hundred grams of the soil fraction < 2 mm was retrieved for analysis on IC, OC and SOM.

5.1.2 Field aggregate stability analysis

Aggregate stability was tested in the field with a field soil aggregate stability kit described by Herrick et al. (2001) (further: Herrick test). The aggregates were submerged individually in 1.65 mm sieves in demineralized water. The slaking of the aggregates was observed until 300 seconds passed. After this the sieves were lifted out of the water and dipped back in for five times. The remaining structure of the aggregates was rated following the criteria as shown in Table 1. Aggregate stability was also tested in the laboratory, which is described in the following paragraph.

Table 1. Criteria for the assignment of aggregates to stability classes (Herrick et al., 2001). Stability class Criteria for assignment to stability class

0 Soil too instable to sample (falls through sieve)

1 50 % of structural integrity lost within 5 s of insertion in water

2 50 % of structural integrity lost within 5 - 30 s of insertion in water

3 50 % of structural integrity lost within 30 - 300 s of insertion in water or < 10 % of soil remains on sieve after five dipping cycles

4 10 - 25 % of soil remains on sieve after five dipping cycles

5 25 - 75 % of soil remains on sieve after five dipping cycles

(8)

8 5.2 Laboratory analysis

In a laboratory environment the samples were analyzed to gain insight in the following characteristics: aggregate stability, SMC, SOM, inorganic carbon content (IC) and OC. To investigate OC and IC the samples < 2 mm were to be milled. This was done by placing a tablespoon of soil into a mortar together with three marbles to act as pestle. Four mortars a time were added to a grinding machine for 5 minutes at 400 rpm. IC and OC of the grinded sample was determined with the Van Wesemael method and CNS analysis (Kroneman, 1971; Elementar Germany, 2017). 5.2.1 Aggregate stability

Before testing aggregate stability in the laboratory the aggregates were moistened with distilled water at pF 1 for 24 hours. Aggregate stability was tested by Counting the Number of water Drops (CND) required to fracture the aggregate enough for it to pass through a 2.8 mm sieve, as described by Imeson & Vis (1984). The water drops fell 1 meter down through a 15 cm diameter polythene pipe before impacting the aggregates. This CND test was repeated twenty times per sample.

5.2.2 Soil organic matter

The SOM determination was based on the method of Loss on Ignition (LOI) (Dean, 1974). This method requires the SMC to be removed first. The SMC was removed by drying 4 – 5 grams of the < 2 mm soil fraction overnight at 105 °C. The samples were weighed before and after drying. The difference in weight can be used to calculate the SMC of the soil as shown in equation 1.

[1] 𝑆𝑀𝐶 [%] = 𝑂𝑟𝑖𝑔𝑖𝑛𝑎𝑙 𝑠𝑎𝑚𝑝𝑙𝑒 𝑤𝑒𝑖𝑔ℎ𝑡 [𝑚𝑔] − 𝐷𝑟𝑖𝑒𝑑 𝑠𝑎𝑚𝑝𝑙𝑒 𝑤𝑒𝑖𝑔ℎ𝑡 [𝑚𝑔]_{𝑂𝑟𝑖𝑔𝑖𝑛𝑎𝑙 𝑠𝑎𝑚𝑝𝑙𝑒 𝑤𝑒𝑖𝑔ℎ𝑡 [𝑚𝑔]} · 100%

SOM was determined by placing the dried soil samples in an oven at 500 °C overnight. This causes the organic matter to combust. The samples were weighed when they were retrieved from the oven. The difference in weight with the dried samples is used to calculate SOM as shown in equation 2.

[2] 𝑆𝑂𝑀 [%] = 𝐷𝑟𝑖𝑒𝑑 𝑠𝑎𝑚𝑝𝑙𝑒 𝑤𝑒𝑖𝑔ℎ𝑡 [𝑚𝑔] − 𝐼𝑔𝑛𝑖𝑡𝑒𝑑 𝑠𝑎𝑚𝑝𝑙𝑒 𝑤𝑒𝑖𝑔ℎ𝑡 [𝑚𝑔]

𝐷𝑟𝑖𝑒𝑑 𝑠𝑎𝑚𝑝𝑙𝑒 𝑤𝑒𝑖𝑔ℎ𝑡 [𝑚𝑔] · 100%

5.2.3 Inorganic carbon content

IC was determined by using the Van Wesemael method (Kroneman, 1971). This method uses the gravimetric loss of CaCO3 as a proxy to determine the carbonate content in the soil. The loss of

carbonates is caused by a reaction of CaCO3 with HCl resulting in a production of CO2. This reaction

is shown in equation 3.

[3] 𝐶𝑎𝐶𝑂3+ 𝐻+→ 𝐶𝑎2++ 𝐶𝑂2+ 𝐻2𝑂

Four to five grams of air-dried soil was added to a glass container. To this container a glass tube filled with a 4 M HCl solution was added. The total was weighed and the container was tilted to allow the HCl solution to react with the soil. The container was placed into a shaker, gently shaking every 30 minutes for a 24 hour period. After this the containers were weighed again to determine the gravimetric loss. To save laboratory time, only 18 samples were used to determine IC. IC of other samples were calculated using statistical analysis, which will be explained in the next

(9)

9 paragraph. To calculate IC the evaporated CO2 content was calculated using equation 4. The

outcome of equation 4 was used in equation 5 to calculate the content of CaCO3 and thus IC.

[4] 𝐶𝑂₂[%] = 𝑃 𝑥 𝑔𝑟𝑎𝑚𝑠 𝐶𝑎𝐶𝑂3 𝑥 44

𝑄 𝑥 100 𝑥 𝑅 · 100%

P: gravimetric loss of the sample in grams Q: gravimetric loss of CaCO3 in grams

R: grams of air dry sample

[5] 𝐶𝑎𝐶𝑂3[%] = 𝐶𝑂2[%] ·

100 44

5.2.4 Organic carbon content

OC was determined by measuring the total carbon content (TC) present in the soil. By subtracting IC from TC, OC can be calculated.

Approximately 50 mg of the milled soil was added to an aluminum foil cup. The cup was closed and pressed into a disk-shaped container. In addition, measuring standards of 5 – 15 mg of sulfanilic acid were folded by hand. The packages were placed into a Vario EL Cube Elemental Analyser by Elementar Analyse Systems. This is an automatic element analyzer, that automatically loads individual samples and measures element levels through flash combustion at 1800 °C with an excess of oxygen (Elementar Germany, 2017). The products of the combustion are removed from the combustion chamber by He gas and moved over heated high purity copper, removing the remaining free O2 from the combustion. After passing through more traps in the selective trap

columns N2, CO2, H2O and SO2 are the only particles left. The masses of these particles were used

to calculate the TC, total nitrogen content and total sulfur content. This process is shown in Figure 3.

(10)

10 5.3 Statistical analysis

Data processing and structuring was done using Excel 2013 as preparation for statistical analysis. Statistical analysis and data visualization was carried out using Matlab R2016b.

First, OC was calculated by subtracting IC from TC. SOM and OC were analyzed using simple regression in order to predict the missing OC and IC values that were not retrieved from the laboratory. Some values of the predicted OC exceeded the measured TC. Because this is not possible, the fitted OC was capped off at the TC. Therefore, the assumption was made that in those samples no IC was present.

ANOVA analysis could not be done to examine the distributions and the variances of the results because the results were not distributed normally. To overcome this a Wilcoxon signed rank-sum test was used to analyze the distributions of the different sample locations in more detail. This statistical analysis tests the following hypotheses:

H0: Data in x and y are samples from continuous distributions with equal medians H1: Data in x and y are not samples from continuous distributions with equal medians

The Wilcoxon rank-sum test was also used to analyze the data from the hillslope versus alluvial fan and Macrochloa tenacissima versus Anthyllis cytisoides.

A Kruskal-Wallis test was used to examine data divided into more specific groups. This statistical analysis tests the following hypotheses:

H0: All samples come from the same distribution H1: Not all samples come from the same distribution The data that was examined is shown in Table 2.

Table 2. The different results that were tested against each other. B: Bare, C: Canopy border S: Stem, F: Fan, H: Hill, A:

Anthyllis cytisoides, M: Macrochloa tenacissima.

Kruskal-Wallis test 1 Kruskal-Wallis test 2 Kruskal-Wallis test 4 Kruskal-Wallis test 3 Kruskal-Wallis test 5 B BF AB AF ABF C BH AC AH ACF S CF AS MF ASF CH MB MH ABH SF MC ACH SH MS ASH MBF MCF MSF MBH MCH MSH

The Wilcoxon rank-sum test and Kruskal-Wallis test as described were used for the following parameters: SOM, TC, IC and CND. For CND all individual measurements were used in the analysis. Simple regression was used to examine the relation between parameters. This was done for SOM versus the median of CND, SOM versus Herrick test, SOM versus TC and SOM versus OC.

(11)

11

6. Results

6.1 Regression analysis

The results of the performed regression analyses are shown in Figure 4. Tables with raw data can be found in appendix A.3.

Figure 4. Regression analyses performed. The regression equation, R2_{and p-values are shown in the regression plots.}

Below that the plots of the standardized residuals are shown.

Figure 4 shows that no positive relationship between SOM and CND was found. The R2_{of 0.01}

indicates an explained variance of 1%. This means that the model does not explain the variance in the model well. The p-value of 0.47 shows that the slope of the linear model does not differ significantly from zero, when using α = 0.05. A Spearman rank correlation analysis returned a p-value of 0.1522, showing no correlation between the two variables. Additionally, regression analysis with a selection of the data (e.g. only data from locations of alluvial fan) did not result in a significant positive relation between SOM and aggregate stability.

The second calculated model for prediction of TC with the independent variable SOM returned better result. The R2_{of the model showed that 71% of the variance of TC is explained by the model.}

The slope of the line differs significantly from 0 as indicated by the p-value < 0.001, indicating a positive relation between the two variables.

The third model for predicting OC is even a better fit than the second model. The R2_{of 0.94}

indicates that almost all variance of the response variable OC is explained. The p-value of < 0.001 for the slope of the model indicates a significant difference from 0, and thus indicating a significant positive relation between the two variables.

The three bottom plots show the residuals of the three fitted models. The plots show that no assumptions of linear regression are violated. Hence, the results of these regression analyses are valid. The regression analysis for SOM and Herrick test did not return a well-fitted positive linear relation (Figure 8 in appendix A.2).

(12)

12 6.2 Boxplots

The results of the performed laboratory analyses are shown in Figure 5-7 by means of boxplots. P-values for significant pairs of results are shown in Table 3. However, not all significant pairs are shown as for some pairs two or more indices between sample locations differ. These pairs do not lead to useful conclusions and are therefore not touched upon. They can be found in Table 4 in appendix A.1. Tables including raw data can be found in appendix A.3.

Figure 5. Boxplots of laboratory data. Horizontal red lines indicate the median, blue boxes indicate the interquartile range, red crosses outside whiskers indicate outliers. B: Bare, C: Canopy border S: Stem, F: Fan, H: Hill. B-C-S indicate clusters of all results for samples positioned at respectively bare, canopy border and stem locations. H-F indicate clusters of all results for samples positioned at the hillside and the alluvial fan respectively. BF-BH-CF-CH-SF-SH indicate clusters of all results for samples positioned on bare spots in the alluvial fan, bare spots on the hillslope etc. P-values for all significant pairs are shown in Table 3.

Figure 5 shows the distribution of the gained laboratory data for the clusters results Bare – Canopy border – Stem, Alluvial fan – Hillslope and a combinations of those indices and is thus not taking vegetation species into account.

Similar patterns can be seen for SOM and TC: the content is highest for results from samples near the stem and lowest for results located on bare patches. The Wilcoxon rank-sum test shows that SOM at bare patches is significantly lower than near the stem of vegetation. SOM and TC are significantly higher for results positioned on the hillslope than on the alluvial fan. The results from sample locations at the canopy border on the hillslope are significantly higher than those on the alluvial fan.

(13)

13 For IC no obvious pattern is visible. For CND no difference is seen between aggregate stability on the hillslope versus the alluvial fan. A slight upward trend is visible when comparing results from Bare - Canopy border – Stem locations. Table 3 shows that all three sample locations differ significantly. Aggregate stability at locations on bare patches on hillslope is significantly lower than on locations near the stem of vegetation on hillslope. This is the similar for bare and stem locations positioned on the alluvial fan. For all variables, outliers can be observed. However, IC shows most outliers.

Figure 6. Boxplots of laboratory data. Horizontal red lines indicate the median, blue boxes indicate the interquartile range, red crosses outside whiskers indicate outliers. A: Anthyllis cytisoides, M: Macrochloa tenacissima, B: Bare, C: Canopy border S: Stem. A-M indicate clusters of all results for samples positioned at respectively bare, canopy border and stem locations. AB-AC-AS-MB-MC-MS indicate clusters of all results for samples positioned on bare spots near Anthyllis cytisoides, on the canopy border of an Anthyllis plant etc. P-values for all significant pairs are shown in Table 3.

Figure 6 shows the distribution of the gained laboratory data for the clusters results Anthyllis

cytisoides – Macrochloa tenacissima and partitioning of those clusters based on the sample

location with respect to the vegetation patch. For all parameters, the clustered results for Anthyllis

cytisoides and Macrochloa tenacissima did not differ significantly. However, the aggregate stability

near the stem of the Macrochloa tenacissima was significantly higher than those near the canopy border and at a bare patch near the Macrochloa tenacissima.

(14)

14 Figure 7. Boxplots of laboratory data. Horizontal red lines indicate the median, blue boxes indicate the interquartile range, red crosses outside whiskers indicate outliers. A: Anthyllis cytisoides, M: Macrochloa tenacissima, F: Alluvial fan, H: Hillslope, B: Bare, C: Canopy border S: Stem. AF-AH-MF-MH indicate clusters of all results for samples positioned at Anthyllis cytisoides on the alluvial fan, Anthyllis cytisoides on the hillslope etc. ABF-ACF-ASF-ACH-etc. indicate clusters of all results for samples positioned on bare spots near Anthyllis cytisoides on the alluvial fan, on the canopy border of an Anthyllis plant on the alluvial fan etc. P-values for all significant pairs are shown in Table 3.

Figure 7 shows the distribution of the gained laboratory data for the clusters results Anthyllis

cytisoides – Macrochloa tenacissima and partitioning of those clusters based on the sample

location with respect to the vegetation patch. Both Anthyllis cytisoides and Macrochloa tenacissima show higher SOM and TC at the hillslope than on the alluvial fan. However, only at the Macrochloa

tenacissima a significant difference was found.

For CND the only two related groups that differ significantly are MBF and MSF. This indicating that the aggregate stability on the alluvial fan is higher near the stem of the Macrochloa tenacissima when compared to bare soil.

(15)

15 6.3 Wilcoxon rank-sum and Kruskal-Wallis analysis

The significant p-values of the Wilcoxon rank sum test and the Kruskal-Wallis test are shown in Table 3 for the variables SOM, TC and CND. No significant results were found for IC.

Not all significant pairs are shown as for some pairs two or more indices between sample locations differ. These pairs do not lead to useful conclusions and are therefore not touched upon.

Table 3. Pairs of results that differ significantly and have only one different index. All shown pairs but H-F are analyzed using a Kruskal-Wallis test. H-F is analyzed using a Wilcoxon rank-sum test. B: Bare, C: Canopy border S: Stem, F: Fan, H: Hill, A: Anthyllis cytisoides, M: Macrochloa tenacissima, WRS: Wilcoxon rank-sum, KW1-5: Kruskal-Wallis test 1-5 as indicated in Table 2.

Significant pair p-value SOM p-value TC p-value CND Test

H F 0.003 < 0.001 WRS B C 0.017 KW1 S C 0.031 KW1 B S 0.009 < 0.001 KW1 BF SF < 0.001 KW2 BH SH 0.017 KW2 CF CH 0.035 KW2 SF SH 0.030 KW2 MF MH 0.019 0.002 KW3 MB MS < 0.001 KW4 MC MS 0.040 KW4 MBF MSF 0.004 KW5

(16)

16

7. Discussion

This research studied the impact of Anthyllis cytisoides and Macrochloa tenacissima on soil structural stability in the Rambla Honda Basin in southeastern Spain. Samples were retrieved near both Anthyllis cytisoides and Macrochloa tenacissima on the hillslope and the alluvial fan, while also incorporating the position with respect to the vegetation patch.

The results of the regression analysis between SOM and CND were not in line with the expectations. According to literature (Cerdà, 1998; Cantón et al., 2009), higher aggregate stability should occur when SOM is higher. However, when the data is observed, only a slightly positive relation can be found (Figure 4, top left). The variance of CND explained by the model is only 1 %, which indicates that the model is not a good fit. Also, the p-value of the slope is > 0.05, which indicates that the slope does not differ significantly from 0. The relation was also analyzed for selections of the data (e.g. only samples located on the alluvial fan or near Macrochloa

tenacissima). This also did not return significant positive relationships that explained enough of

the variance of aggregate stability. When analyzing the relation between SOM and the results from the Herrick field test no significant positive linear relation is found, for all data and selections of it. Therefore the hypotheses that ‘SOM has a positive influence on aggregate stability’ is rejected. Possible explanations for this result may be related to the analysis of the aggregate stability. The water drops of the CND test might have had a kinetic energy that was too high for the retrieved aggregates, as many of the aggregates showed slaking when less than five drops had impacted. The Herrick test shows a lot of maximum scores of 6 for a large amount of samples. This shows that the results from both methods are not in line with each other, because the CND results show poor aggregate stability. First, causes for this discrepancy may be related to possibly damaging the aggregates during transport. Second, biota and clay content can also influence aggregate stability (Bronick and Lal, 2005). These parameters were not measured in this research, thus the relation between these factors and aggregate stability were lack in the analysis. Finally, sampling and measuring errors during research can also be the reason for a different relation between SOM and aggregate stability.

This study showed that SOM at bare patches is significantly lower than near the stem of vegetation, when looking at both vegetation species. The CND results show a slight upward trend when comparing results from Bare - Canopy border – Stem locations and Table 3 shows that all three sample locations differ significantly. These results are in line with literature, that states that vegetation has a positive influence on SOM and soil aggregation (Cammeraat and Imeson, 1998; van Hall et al., 2017). Prior research has found no significant differences for SOM and soil aggregation between Anthyllis cytisoides and Macrochloa tenacissima in the Rambla Honda Basin (Cantón et al., 2009). Results presented in this study support those findings. When analyzing

Anthyllis cytisoides and Macrochloa tenacissima individually, no significant differences for SOM

were found between sampling locations with respect to the vegetation patch. However, an upward trend is visible for both vegetation species. Macrochloa tenacissima shows significant results for aggregate stability between locations near the stem and those near the canopy border and at a bare patch near the vegetation patch. With this information, the hypotheses that SOM and aggregate stability are positively influenced by Macrochloa tenacissima and Anthyllis cytisoides can be confirmed.

The Wilcoxon rank-sum test showed that SOM was significantly higher on the hillslope than on the alluvial fan. This result is in line with literature stating that SOM in the Rambla Honda Basin increases upslope (Nicolau et al., 1996; Puigdefábregas et al., 1996). This is caused by the dominance of Macrochloa tenacissima on the hillslope. Macrochloa tenacissima produces more

(17)

17 plant litter than Anthyllis cytisoides, which is dominant on the alluvial fan, resulting in a larger input of organic matter into the soil and thus a larger SOM on the hillslope. When analyzing

Anthyllis cytisoides and Macrochloa tenacissima individually, samples near Macrochloa tenacissima

on the hillslope are significantly higher than on the alluvial fan. The high density of the Macrochloa

tenacissima plants form a barrier which enhances sedimentation and causes local SOM

accumulation in the soil (Bochet, Rubio and Poesen, 1999). The shallow rooting system of

Macrochloa tenacissima also causes the difference to occur. The soil on the alluvial fan is deeper

than on the hillslope, and the infiltration rates are higher (Puigdefábregas et al., 1996). This prevents the roots of Macrochloa tenacissima from taking up water, as precipitation will infiltrate outside the range of the rooting system. It is interesting to see that SOM near Anthyllis cytisoides on the hillslope and the alluvial fan do not differ significantly. The roots of Anthyllis cytisoides grow deeper and vertically oriented, allowing access to groundwater (Cerdà, 1998) and making it more suitable for the alluvial fan. This causes the Anthyllis cytisoides to stand out more as a resource island and have higher SOM in the soil surrounding it. The significantly higher aggregate stability near the stem of Macrochloa tenacissima on the alluvial fan versus a bare patch is in line with literature and explains that vegetation patches are a resource island in a more bare landscape (Bochet, Rubio and Poesen, 1999). The fact that aggregate stability at bare patches is significantly lower than near the stem of vegetation on both the hillslope and the alluvial fan is also in line with these findings. With this information, the hypothesis that SOM and aggregate stability are higher at vegetation patches on the hillslope than on the alluvial fan can be confirmed.

The results show no significant difference between SOM and aggregate stability of Anthyllis

cytisoides and Macrochloa tenacissima, for all data or selections of it. This is in line with literature

(Bochet, Rubio and Poesen, 1999; Cantón et al., 2009) and leads to confirmation of the hypotheses that Anthyllis cytisoides and Macrochloa tenacissima have similar impact on SOM and aggregate stability.

TC shows overall similar results to SOM. IC is low in all samples, which is characteristic for the Rambla Honda Basin and shows no significant differences between any categories.

The results from this research show that Anthyllis cytisoides and Macrochloa tenacissima have a positive influence on soil structural stability. These findings support the research of Bochet (2006) that found that Anthyllis cytisoides and Macrochloa tenacissima decrease runoff and thus soil erosion. The influence of Macrochloa tenacissima on the hillslope is significantly different from the alluvial fan, the influence of Anthyllis cytisoides is not. After land abandonment bare soil does not allow infiltration of precipitation during irregular rainfall events because surface crusts form an impermeable layer on top of the soil (Contreras, Cantón and Solé-Benet, 2008). Therefore, it is suggested that after abandonment of agricultural fields, replanting with Anthyllis cytisoides on the alluvial fan and Macrochloa tenacissima on the hillslope will reduce soil erosion risk.

Several improvements for future research are suggested. Firstly, analyzing more soil properties (e.g. biota and clay content) may gain more insight into the cause of soil aggregation. Secondly, the results of the CND analysis and Herrick test for analyzing the aggregate stability differed, so further research should use only one of these or other methods. Also, more time in the laboratory would have resulted in a better IC determination because then the values would not have been predicted but measured. Lastly, when more samples are retrieved from the field the data will be more likely to match the assumptions for parametric statistical analysis.

(18)

18

8. Conclusion

This research studied the impact of Anthyllis cytisoides and Macrochloa tenacissima on soil structural stability in the Rambla Honda Basin in southeastern Spain. Samples were retrieved near both Anthyllis cytisoides and Macrochloa tenacissima on the hillslope and the alluvial fan, while also incorporating the position with respect to the vegetation patch. Samples were analyzed on soil organic matter, inorganic carbon, total carbon and aggregate stability in a laboratory environment. When pooled together Anthyllis cytisoides and Macrochloa tenacissima have a positive influence on soil organic matter and aggregate stability. The influence differs for the two locations along the catena. Macrochloa tenacissima has a greater impact on soil organic matter and aggregate stability on the hillslope than on the alluvial fan. The influence of Anthyllis cytisoides does not differ significantly between the hillslope and the alluvial fan. The total carbon shows overall similar results to soil organic matter. Inorganic carbon is low for all samples and shows no significant differences between any sample locations. With this information the research question can be answered: What is the impact of Anthyllis cytisoides and Macrochloa tenacissima on soil

structural stability in the Rambla Honda Basin?

Anthyllis cytisoides and Macrochloa tenacissima have a positive influence on soil structural

stability. This research suggests that after abandonment of agricultural fields replanting with

Anthyllis cytisoides on the alluvial fan and Macrochloa tenacissima on the hillslope will reduce soil

(19)

19

9. Acknowledgements

First and foremost, I would like to thank dr. L.H. Cammeraat for his supervisory and comments during the research project. This helped me shape the research and thesis to become a valuable and educational project. I would like to thank J. Zethof MSc for his supervisory during the fieldwork and for his comments on the thesis. I would like to thank R. L. van Hall MSc for his supervisory in the laboratory and comments on the statistical analysis. Lastly, I would like to thank A.B. Dekkers, A.A. Isarin, E.V. de Jong and M. Wadman for their support during the project.

(20)

20

10. References

Albaladejo, J. et al. (1998) ‘Soil degradation and desertification induced by vegetation removal in a semiarid environment’, Soil Use and Management. Wiley/Blackwell (10.1111), 14(1), pp. 1–5. doi: 10.1111/j.1475-2743.1998.tb00602.x.

Bochet, E., Poesen, J. and Rubio, J. L. (2006) ‘Runoff and soil loss under individual plants of a semi-arid Mediterranean shrubland: influence of plant morphology and rainfall intensity’, Earth

Surface Processes and Landforms. Wiley-Blackwell, 31(5), pp. 536–549. doi: 10.1002/esp.1351.

Bochet, E., Rubio, J. L. and Poesen, J. (1999) ‘Modified topsoil islands within patchy Mediterranean vegetation in SE Spain’, Catena, 38, pp. 23–44. Available at:

www.elsevier.comrlocatercatena (Accessed: 20 June 2018).

Boix, C. et al. (1995) ‘Climatic and altitudinal effects on soil aggregation in slopes of

mediterranean environment’, Physics and Chemistry of the Earth. Pergamon, 20(3–4), pp. 287– 291. doi: 10.1016/0079-1946(95)00039-9.

Bronick, C. J. and Lal, R. (2005) ‘Soil structure and management: a review’, Geoderma. Elsevier, 124(1–2), pp. 3–22. doi: 10.1016/J.GEODERMA.2004.03.005.

Cammeraat, L. H. (2004) ‘Scale dependent thresholds in hydrological and erosion response of a semi-arid catchment in southeast Spain’, Agriculture, Ecosystems & Environment. Elsevier, 104(2), pp. 317–332. doi: 10.1016/J.AGEE.2004.01.032.

Cammeraat, L. H. and Imeson, A. C. (1998) ‘Deriving indicators of soil degradation from soil aggregation studies in southeastern Spain and southern France’, Geomorphology, 23(2–4), pp. 307–321.

Cammeraat, L. H. and Imeson, A. C. (1999) ‘The evolution and significance of soil–vegetation patterns following land abandonment and fire in Spain’, Catena, 37, pp. 107–127.

Cantón, Y. et al. (2009) ‘Aggregate stability in range sandy loam soils Relationships with runoff and erosion’, Catena, 77(3), pp. 192–199. doi: 10.1016/j.catena.2008.12.011.

Cerdà, A. (1997) ‘Soil erosion after land abandonment in a semiarid environment of

southeastern Spain’, Arid Soil Research and Rehabilitation. Taylor & Francis Group, 11(2), pp. 163–176. doi: 10.1080/15324989709381469.

Cerdà, A. (1998) ‘Soil aggregate stability under different Mediterranean vegetation types’,

Catena, 32, pp. 73–86. doi: 10.1016/S0341-8162(98)00041-1.

Contreras, S., Cantón, Y. and Solé-Benet, A. (2008) ‘Sieving crusts and macrofaunal activity control soil water repellency in semiarid environments: Evidences from SE Spain’. doi: 10.1016/j.geoderma.2008.03.019.

Dean, W. E. (1974) ‘Determination of Carbonate and Organic Matter in Calcareous Sediments and Sedimentary Rocks by Loss on Ignition: Comparison With Other Methods’, SEPM Journal of

Sedimentary Research. GeoScienceWorld, Vol. 44(1), pp. 242–248. doi:

10.1306/74D729D2-2B21-11D7-8648000102C1865D.

Elementar Germany (2017) ‘vario EL Cube Elemental Analyser’. Available at:

https://www.elementar.de/fileadmin/user_upload/Elementar_Website/Downloads/Flyer/Flye r-vario-EL-cube-EN.pdf (Accessed: 7 June 2018).

van Hall, R. L. et al. (2017) ‘Impact of secondary vegetation succession on soil quality in a humid Mediterranean landscape’, Catena. Elsevier B.V., 149, pp. 836–843. doi:

(21)

21 Herrick, J. E. et al. (2001) ‘Field soil aggregate stability kit for soil quality and rangeland health evaluations’, Catena, 44(1), pp. 27–35. doi: 10.1016/S0341-8162(00)00173-9.

Imeson, A. C. and Vis, M. (1984) ‘Assessing soil aggregate stability by water-drop impact and ultrasonic dispersion’, Geoderma. Elsevier, 34(3–4), pp. 185–200. doi:

10.1016/0016-7061(84)90038-7.

Kroneman, H. (1971) Bepaling van het calciumcarbonaatgehalte volgens de methode Van

Wesemael. Kampen.

Martínez-Mena, M. et al. (2002) ‘Organic carbon and nitrogen losses influenced by vegetation removal in a semiarid mediterranean soil’, Biogeochemistry. Kluwer Academic Publishers, 61(3), pp. 309–321. doi: 10.1023/A:1020257208048.

Nicolau, J. M. et al. (1996) ‘Effects of soil and vegetation on runoff along a catena in semi-arid Spain’, Geomorphology. Elsevier, 14(4), pp. 297–309. doi: 10.1016/0169-555X(95)00043-5. Norton, J. B., Sandor, J. A. and White, C. S. (2003) ‘Hillslope Soils and Organic Matter Dynamics within a Native American Agroecosystem on the Colorado Plateau’, Soil Science Society of

America Journal, 67(1), pp. 225–234.

Oztas, T., Koc, A. and Comakli, B. (2003) ‘Changes in vegetation and soil properties along a slope on overgrazed and eroded rangelands’, Journal of Arid Environments, 55(1), pp. 93–100. doi: 10.1016/S0140-1963(02)00267-7.

Puigdefábregas, J. et al. (1996) ‘The Rambla Honda filed site: Interactions of soil and vegetation along a catena in semi-arid Southeast Spain’, Mediterranean Desertification and Land Use, (June), pp. 137–168. doi: 201.

Puigdefábregas, J. et al. (1998) ‘Atlasof Mediterranean Environments in Europe’, in Atlas of

(22)

22

A. Appendix

A.1. Significant pairs of results

Table 4. Pairs of results that differ significantly, and have two or more different indices. All shown pairs are analyzed using a Kruskal-Wallis test. B: Bare, C: Canopy border S: Stem, F: Fan, H: Hill, A: Anthyllis cytisoides, M: Macrochloa

tenacissima, WRS: Wilcoxon rank-sum, KW1: Kruskal-Wallis test 1 as indicated in Table 2.

Significant pair p-value SOM p-value TC p-value CND Test

BF CH 0.043 KW2 BF SH 0.020 0.003 KW2 BH SF 0.002 KW2 CF SH 0.017 0.002 KW2 MF AH 0.006 0.001 KW3 MB AC 0.012 KW4 MB AS 0.035 0.016 KW4 MS AB 0.006 KW4 MBF ACF 0.040 KW5 MBF ASH 0.028 KW5 MCF MSH 0.029 KW5 MCF ASH 0.049 KW5

(23)

23 A.2. Regression analysis SOM and Herrick test

(24)

24 A.3. Raw data

A.3.1. Soil moisture content and soil organic matter

Table 5. SMC and SOM. B: Bare, C: Canopy border S: Stem, F: Fan, H: Hill, A: Anthyllis cytisoides, M: Macrochloa

tenacissima, 1-5: field transact 1-5.

Sample SMC [%] SOM [%] Sample SMC [%] SOM [%] Sample SMC [%] SOM [%]

1FAB2 5.84384 4.48036 2HAS 1.32008 5.27710 4FMC2 1.95295 1.63401 1FAB2 5.89593 4.48336 2HAS 1.40532 5.42551 4FMC2 2.18777 2.57462 1FAC2 8.25644 3.23100 2HMB2 0.38282 2.83115 4FMS 2.90204 3.51944 1FAC2 8.15043 3.94385 2HMB2 0.39588 3.38238 4FMS 2.74857 3.67369 1FAS 4.36726 5.37823 2HMC2 0.82057 7.01092 4HAB2 1.64637 2.18348 1FAS 4.43020 5.72544 2HMC2 0.88554 7.69299 4HAB2 1.75050 2.29074 1FMB2 6.80038 5.21150 2HMS 1.93037 7.19013 4HAC2 8.15676 7.67793 1FMB2 7.17270 5.65877 2HMS 2.05035 7.28946 4HAC2 8.33006 8.07081 1FMC2 9.50725 3.80495 3FAB2 0.65424 2.10377 4HAS 8.65909 5.69139 1FMC2 9.38865 5.19561 3FAB2 0.67341 2.26352 4HAS 8.11574 6.64579 1FMS 10.17873 5.23083 3FAC2 1.12301 3.97821 4HMB2 2.00945 2.53077 1FMS 9.98670 5.36268 3FAC2 1.18449 3.99945 4HMB2 2.09217 2.64256 1HAB2 0.75576 3.58456 3FAS 3.12354 4.85213 4HMC2 1.68822 2.77154 1HAB2 0.71936 3.65230 3FAS 3.11830 4.96405 4HMC2 1.65889 2.79708 1HAC2 0.81331 5.81362 3FMB2 2.50260 2.34354 4HMS 6.77889 4.91307 1HAC2 1.05213 6.61572 3FMB2 2.35622 2.83504 4HMS 6.84519 5.53504 1HAS 2.45915 12.87726 3FMC2 3.85482 1.84291 5FAB2 1.15835 2.77274 1HAS 2.82798 14.85608 3FMC2 3.93778 2.01752 5FAB2 1.19991 3.10663 1HMB2 0.73299 4.06330 3FMS 3.97177 3.08565 5FAC2 1.89525 3.47265 1HMB2 0.58788 4.26845 3FMS 4.02003 3.22027 5FAC2 1.95105 4.17955 1HMC2 1.43927 9.03181 3HAB2 5.63134 4.42384 5FAS 5.80555 7.40288 1HMC2 1.48188 9.41870 3HAB2 5.75260 5.11136 5FAS 6.70633 8.71431 1HMS 1.31321 4.55394 3HAC2 4.31538 1.91828 5FMB2 1.86950 3.34322 1HMS 1.13190 5.20695 3HAC2 4.37023 2.07175 5FMB2 1.86218 3.61312 2FAB2 6.55872 4.08484 3HAS 4.63826 3.06482 5FMC2 2.48467 1.58427 2FAB2 6.78773 4.79365 3HAS 4.64167 3.24234 5FMC2 2.48897 2.34509 2FAC2 8.40664 6.24734 3HMB2 3.34029 1.84617 5FMS 2.13235 2.01854 2FAC2 8.28907 6.58068 3HMB2 2.99444 2.54510 5FMS 2.06092 2.52039 2FAS 8.05881 6.83629 3HMC2 6.64264 5.40831 5HAB2 1.50530 10.49331 2FAS 8.41192 7.52583 3HMC2 6.52385 6.74370 5HAB2 1.62854 11.56913 2FMB2 6.00332 2.01107 3HMS 8.05103 5.85293 5HAC2 4.77913 8.49764 2FMB2 5.99255 2.03387 3HMS 8.63575 7.38715 5HAC2 5.10336 9.01676 2FMC2 5.16112 2.22814 4FAB2 0.69589 2.70517 5HAS 3.79383 7.53050 2FMC2 5.18751 2.32984 4FAB2 0.69118 2.91560 5HAS 3.99928 7.79401 2FMS 5.17969 2.87142 4FAC2 3.98220 1.95227 5HMB2 1.05263 4.12103 2FMS 5.05946 3.38416 4FAC2 4.22843 2.64406 5HMB2 1.12254 NaN 2HAB2 0.52771 2.22213 4FAS 4.52660 3.30722 5HMC2 4.46853 6.05594 2HAB2 0.56257 2.41453 4FAS 4.46057 3.46338 5HMC2 4.36032 6.38688 2HAC2 1.31170 7.19142 4FMB2 1.15645 2.07907 5HMS 8.13923 7.24650 2HAC2 1.39779 7.61718 4FMB2 1.08210 2.43719 5HMS 9.38576 8.29655

(25)

25 A.3.2. Total carbon, nitrogen and sulfur

Table 6. Total carbon, nitrogen and sulfur. B: Bare, C: Canopy border S: Stem, F: Fan, H: Hill, A: Anthyllis cytisoides, M:

Macrochloa tenacissima, 1-5: field transact 1-5.

Sample C [%] N [%] S [%] Sample C [%] N [%] S [%] Sample C [%] N [%] S [%] 1FAB2 2.35 0.26 0.06 2HAS 3.17 0.26 0.04 4FMC2 0.98 0.11 0.03 1FAB2 2.36 0.26 0.05 2HAS 3.25 0.26 0.04 4FMC2 0.98 0.12 0.03 1FAC2 1.82 0.19 0.04 2HMB2 1.28 0.15 0.06 4FMS 1.42 0.15 0.03 1FAC2 1.90 0.20 0.04 2HMB2 1.31 0.15 0.05 4FMS 1.38 0.15 0.04 1FAS 3.00 0.27 0.04 2HMC2 3.94 0.35 0.06 4HAB2 1.98 0.20 0.04 1FAS 2.86 0.25 0.05 2HMC2 3.96 0.35 0.05 4HAB2 2.24 0.22 0.04 1FMB2 3.08 0.27 0.04 2HMS 3.62 0.32 0.05 4HAC2 3.69 0.34 0.04 1FMB2 2.98 0.26 0.05 2HMS 3.55 0.31 0.05 4HAC2 3.49 0.32 0.05 1FMC2 2.38 0.22 0.04 3FAB2 3.17 0.30 0.05 4HAS 3.14 0.30 0.04 1FMC2 2.48 0.23 0.05 3FAB2 2.52 0.24 0.04 4HAS 3.13 0.30 0.04 1FMS 0.77 0.10 0.03 3FAC2 1.56 0.17 0.04 4HMB2 3.01 0.28 0.04 1FMS 0.61 0.08 0.02 3FAC2 1.59 0.16 0.04 4HMB2 3.61 0.32 0.04 1HAB2 1.82 0.19 0.04 3FAS 2.07 0.20 0.05 4HMC2 1.13 0.14 0.03 1HAB2 1.80 0.19 0.04 3FAS 2.03 0.20 0.04 4HMC2 1.14 0.14 0.02 1HAC2 1.44 0.16 0.03 3FMB2 1.24 0.13 0.03 4HMS 3.79 0.33 0.05 1HAC2 1.64 0.18 0.04 3FMB2 1.23 0.13 0.03 4HMS 4.40 0.38 0.05 1HAS 7.89 0.57 0.09 3FMC2 0.83 0.11 0.03 5FAB2 1.62 0.17 0.04 1HAS 7.97 0.59 0.08 3FMC2 0.85 0.11 0.03 5FAB2 1.60 0.16 0.03 1HMB2 1.66 0.17 0.04 3FMS 1.42 0.15 0.03 5FAC2 2.32 0.20 0.04 1HMB2 1.67 0.17 0.05 3FMS 1.41 0.15 0.03 5FAC2 2.32 0.20 0.04 1HMC2 4.85 0.41 0.08 3HAB2 2.15 0.21 0.05 5FAS 0.88 0.13 0.02 1HMC2 4.68 0.40 0.06 3HAB2 2.14 0.20 0.05 5FAS 1.02 0.15 0.03 1HMS 2.94 0.26 0.06 3HAC2 0.78 0.10 0.03 5FMB2 1.73 0.16 0.03 1HMS 2.79 0.24 0.05 3HAC2 0.78 0.11 0.03 5FMB2 1.77 0.17 0.04 2FAB2 2.12 0.19 0.05 3HAS 1.23 0.13 0.03 5FMC2 0.85 0.11 0.03 2FAB2 2.12 0.20 0.05 3HAS 1.21 0.13 0.03 5FMC2 0.87 0.11 0.03 2FAC2 1.01 0.11 0.03 3HMB2 0.98 0.12 0.04 5FMS 1.17 0.12 0.04 2FAC2 1.19 0.13 0.03 3HMB2 0.99 0.12 0.03 5FMS 1.14 0.12 0.04 2FAS 3.79 0.33 0.07 3HMC2 2.73 0.24 0.04 5HAB2 5.72 0.48 0.07 2FAS 3.97 0.34 0.07 3HMC2 2.80 0.25 0.04 5HAB2 5.86 0.49 0.06 2FMB2 0.80 0.11 0.05 3HMS 3.66 0.30 0.04 5HAC2 4.99 0.44 0.05 2FMB2 0.78 0.11 0.03 3HMS 3.43 0.27 0.04 5HAC2 4.78 0.42 0.05 2FMC2 0.71 0.10 0.04 4FAB2 1.38 0.14 0.04 5HAS 3.83 0.32 0.05 2FMC2 0.72 0.10 0.03 4FAB2 1.41 0.14 0.04 5HAS 3.79 0.32 0.05 2FMS 1.30 0.15 0.03 4FAC2 1.05 0.12 0.04 5HMB2 1.67 0.17 0.03 2FMS 1.31 0.14 0.04 4FAC2 1.04 0.12 0.04 5HMB2 1.67 0.16 0.03 2HAB2 NaN NaN NaN 4FAS 1.52 0.15 0.04 5HMC2 2.99 0.26 0.04 2HAB2 NaN NaN NaN 4FAS 1.52 0.15 0.04 5HMC2 3.00 0.26 0.04 2HAC2 3.49 0.30 0.06 4FMB2 0.91 0.11 0.03 5HMS 3.97 0.33 0.05 2HAC2 3.57 0.30 0.05 4FMB2 0.89 0.10 0.03 5HMS 4.00 0.34 0.05

(26)

26 A.3.3. Inorganic and organic carbon

Table 7. TC, IC and OC. TC is retrieved from Table 6. OC is calculated by subtracting IC from TC. B: Bare, C: Canopy border S: Stem, F: Fan, H: Hill, A: Anthyllis cytisoides, M: Macrochloa tenacissima, 1-5: field transact 1-5.

Sample TC [%] IC [%] OC [%] 1FAC2 1.8600 0.0053 1.8547 1HMS 2.8650 0.0843 2.7807 2FAS 3.8800 0.0108 3.8692 2FMC2 0.7150 0.0000 0.7150 2FMS 1.3050 0.0083 1.2967

2HAB2 NaN 0.0077 NaN

3FAS 2.0500 0.0105 2.0395 3FMB2 1.2350 0.0532 1.1818 3FMS 1.4150 0.0067 1.4083 3HMB2 0.9850 0.0095 0.9755 4FAB2 1.3950 0.0823 1.3127 4HAC2 3.5900 0.0217 3.5683 5FMB2 1.7500 0.0977 1.6523 5HAB2 5.7900 0.0078 5.7822 5HAS 3.8100 0.0313 3.7787 5HMC2 2.9950 0.0051 2.9899

(27)

27 A.3.4. Aggregate stability: Herrick test

Table 8. Results of Herrick field test. B: Bare, C: Canopy border S: Stem, F: Fan, H: Hill, A: Anthyllis cytisoides, M:

Macrochloa tenacissima, 1-5: field transact 1-5, HS1-HS8: repetition of Herrick test

Sample HS1 HS2 HS3 HS4 HS5 HS6 HS7 HS8 Sample HS1 HS2 HS3 HS4 HS5 HS6 HS7 HS8 1FAB2 3 5 5 6 6 6 6 6 3HAB2 4 5 6 6 6 6 6 6 1FAC2 3 4 6 6 6 6 6 6 3HAC2 4 4 4 4 5 5 5 6 1FAS 3 6 6 6 6 6 6 6 3HAS 5 6 6 6 6 6 6 6 1FMB2 3 3 3 4 5 6 6 6 3HMB2 3 3 3 3 4 5 6 6 1FMC2 5 6 6 6 6 6 6 6 3HMC2 4 5 6 6 6 6 6 6 1FMS 6 6 6 6 6 6 6 6 3HMS 6 6 6 6 6 6 6 6 1HAB2 6 6 6 6 6 6 6 6 4FAB2 5 5 6 6 6 6 6 6 1HAC2 6 6 6 6 6 6 6 6 4FAC2 5 6 6 6 6 6 6 6 1HAS 6 6 6 6 6 6 6 6 4FAS 5 6 6 6 6 6 6 6 1HMB2 5 5 5 6 6 6 6 6 4FMB2 3 3 3 3 4 5 5 6 1HMC2 5 6 6 6 6 6 6 6 4FMC2 3 3 3 3 4 5 6 6 1HMS 5 6 6 6 6 6 6 6 4FMS 4 5 6 6 6 6 6 6 2FAB2 3 6 6 6 6 6 6 6 4HAB2 3 4 5 5 5 6 6 6 2FAC2 6 6 6 6 6 6 6 6 4HAC2 5 6 6 6 6 6 6 6 2FAS 4 5 6 6 6 6 6 6 4HAS 6 6 6 6 6 6 6 6 2FMB2 3 4 5 6 6 6 6 6 4HMB2 5 5 6 6 6 6 6 6 2FMC2 5 5 6 6 6 6 6 6 4HMC2 4 5 6 6 6 6 6 6 2FMS 4 5 6 6 6 6 6 6 4HMS 3 4 5 6 6 6 6 6 2HAB2 4 5 5 6 6 6 6 6 5FAB2 4 5 6 6 6 6 6 6 2HAC2 5 5 6 6 6 6 6 6 5FAC2 5 6 6 6 6 6 6 6 2HAS 6 6 6 6 6 6 6 6 5FAS 5 6 6 6 6 6 6 6 2HMB2 3 3 4 5 6 6 6 6 5FMB2 3 5 6 6 6 6 6 6 2HMC2 5 6 6 6 6 6 6 6 5FMC2 6 6 6 6 6 6 6 6 2HMS 5 6 6 6 6 6 6 6 5FMS 6 6 6 6 6 6 6 6 3FAB2 3 3 4 4 5 5 6 6 5HAB2 5 6 6 6 6 6 6 6 3FAC2 5 6 6 6 6 6 6 6 5HAC2 3 6 6 6 6 6 6 6 3FAS 6 6 6 6 6 6 6 6 5HAS 5 6 6 6 6 6 6 6 3FMB2 3 3 3 4 5 5 6 6 5HMB2 3 3 4 4 4 5 6 6 3FMC2 3 4 5 5 5 5 5 6 5HMC2 4 5 5 6 6 6 6 6 3FMS 3 5 5 6 6 6 6 6 5HMS 5 5 6 6 6 6 6 6

(28)

28 A.3.5. Aggregate stability: CND

Table 9. Results of CND analysis. 1-5: field transact 1-5, B: Bare, C: Canopy border S: Stem, F: Fan, H: Hill, A: Anthyllis

cytisoides, M: Macrochloa tenacissima, 1-20 repetition of CND analysis

Sample 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 1FAB2 2 3 29 6 30 8 17 2 2 2 3 5 3 117 2 3 1 2 2 2 1FAC2 3 6 6 6 24 4 5 6 3 4 14 5 4 3 6 6 5 4 6 5 1FAS 24 3 20 8 5 16 2 6 3 2 4 3 27 3 3 3 3 4 2 5 1FMB2 12 3 3 2 8 3 5 2 15 2 4 4 3 5 3 4 92 5 5 4 1FMC2 3 4 4 5 3 7 5 4 5 6 2 5 3 4 4 3 4 4 7 5 1FMS 8 3 14 5 28 4 9 12 5 9 10 5 5 7 36 21 7 24 10 37 1HAB2 3 4 8 12 21 3 4 4 4 10 5 2 3 6 10 2 2 2 5 3 1HAC2 4 6 5 6 4 5 2 7 4 6 5 4 6 3 6 4 3 4 2 3 1HAS 2 5 2 3 15 4 4 2 9 4 5 3 15 14 7 2 5 2 2 3 1HMB2 5 5 4 6 3 4 4 9 5 6 4 4 5 3 1 9 6 2 3 4 1HMC2 3 3 5 7 5 8 2 4 2 5 3 2 3 4 2 3 6 3 5 1 1HMS 26 12 3 7 8 6 5 3 23 4 5 5 15 5 9 19 13 6 3 3 2FAB2 19 5 4 1 3 4 4 18 3 3 3 6 4 4 4 2 3 5 1 3 2FAC2 7 6 3 23 5 3 2 3 5 2 3 6 3 9 7 3 4 4 4 4 2FAS 10 9 4 3 3 4 11 3 2 1 2 10 9 4 3 3 4 11 3 2 2FMB2 5 3 13 7 17 3 5 4 2 7 3 2 3 3 2 2 4 4 3 1 2FMC2 8 7 9 5 8 3 5 2 2 5 6 2 3 3 6 2 3 2 6 5 2FMS 18 9 5 3 4 11 5 10 14 2 5 5 9 3 3 9 3 2 4 6 2HAB2 2 8 11 7 5 1 5 5 2 2 4 3 2 4 3 3 5 3 3 3 2HAC2 2 4 36 4 3 3 5 4 6 4 10 8 4 15 13 6 12 2 2 3 2HAS 3 7 7 2 14 3 6 6 3 3 3 3 2 7 4 3 5 18 9 4 2HMB2 5 3 4 5 4 4 5 7 4 3 6 3 4 6 4 2 4 2 2 3 2HMC2 4 2 8 6 2 3 3 5 4 5 2 2 1 7 8 6 3 4 2 4 2HMS 3 9 4 4 11 3 3 9 4 5 6 5 4 6 3 3 3 3 5 7 3FAB2 3 3 5 5 4 3 4 6 8 4 4 9 2 3 2 5 3 3 3 5 3FAC2 2 2 4 10 6 10 21 4 3 2 6 9 4 11 2 3 4 2 2 4 3FAS 9 2 3 3 3 8 3 3 5 2 4 6 12 4 5 18 4 8 6 6 3FMB2 3 2 9 4 3 3 3 2 2 7 5 5 3 1 2 2 3 1 2 4 3FMC2 6 6 7 4 3 6 5 2 7 2 3 6 2 3 3 5 2 6 4 8 3FMS 4 12 5 3 2 2 2 3 3 3 2 2 2 5 3 4 3 5 4 6 3HAB2 6 2 5 3 1 1 2 2 3 1 1 6 2 5 3 1 1 2 2 3 3HAC2 4 5 6 3 3 2 6 4 6 5 3 5 2 4 6 3 3 4 4 4 3HAS 23 10 6 8 21 5 6 6 21 3 4 6 2 2 2 4 7 10 5 3 3HMB2 4 7 5 9 3 3 4 7 3 2 4 2 6 3 2 2 4 3 2 4 3HMC2 2 2 7 5 33 10 5 5 4 20 8 3 3 3 3 6 2 6 4 1 3HMS 3 1 2 4 5 42 5 4 5 5 6 3 5 3 8 4 10 1 2 5

(29)

29 Continuation of Table 9. Results of CND analysis. 1-5: field transact 1-5, B: Bare, C: Canopy border S: Stem, F: Fan, H: Hill, A: Anthyllis cytisoides, M: Macrochloa tenacissima, 1-20 repetition of CND analysis

Sample 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 4FAB2 4 2 5 6 4 6 2 2 3 4 2 3 3 21 11 2 15 5 6 2 4FAC2 19 3 3 14 3 5 3 4 3 3 4 3 10 3 3 5 6 2 4 3 4FAS 4 4 4 3 5 1 4 4 5 10 1 1 2 2 6 2 2 3 3 5 4FMB2 2 4 4 3 3 6 2 5 2 4 2 2 3 2 2 4 1 1 2 2 4FMC2 5 4 2 2 2 2 10 2 1 3 2 5 4 2 2 2 2 10 2 1 4FMS 3 4 4 6 3 5 3 4 4 3 1 4 3 4 5 5 5 29 3 4 4HAB2 3 5 1 1 4 2 1 5 3 5 3 5 1 1 4 2 1 5 3 5 4HAC2 4 3 6 3 4 2 3 3 5 2 2 3 3 1 2 3 4 5 2 2 4HAS 6 5 4 5 8 5 6 3 5 4 5 6 4 3 4 8 4 5 3 3 4HMB2 8 9 4 4 8 2 3 2 4 3 11 3 4 2 6 4 3 7 3 5 4HMC2 4 3 13 2 3 5 4 4 5 4 5 3 4 3 3 3 2 2 3 4 4HMS 10 3 5 6 7 3 4 3 4 3 4 4 4 3 9 3 4 4 4 4 5FAB2 3 2 5 7 3 4 4 3 2 3 2 5 7 3 4 4 3 2 3 2 5FAC2 6 4 10 3 5 3 18 3 2 3 2 4 3 4 5 3 4 3 4 4 5FAS 12 4 3 9 5 5 9 5 4 3 2 4 10 1 2 2 3 6 3 3 5FMB2 8 2 4 2 4 4 4 4 4 2 12 3 1 2 2 4 3 26 1 29 5FMC2 3 8 5 7 8 3 2 3 3 2 4 2 11 4 3 7 4 8 4 6 5FMS 8 4 4 4 3 5 12 2 1 12 6 2 2 4 5 2 3 3 2 2 5HAB2 6 3 21 11 2 4 4 2 5 3 3 5 13 4 5 10 6 4 7 3 5HAC2 26 22 10 2 2 9 5 5 3 12 3 2 7 5 2 5 32 2 2 3 5HAS 2 4 2 10 3 8 2 8 2 3 4 5 4 2 2 3 1 1 13 6 5HMB2 5 13 8 2 2 1 1 2 3 1 3 2 2 1 1 1 8 2 2 3 5HMC2 4 3 9 3 2 6 3 2 2 7 1 2 1 3 3 3 3 2 2 3 5HMS 3 7 7 5 2 1 2 2 2 2 4 3 2 6 1 1 2 2 1 1

The impact of Macrochloa tenacissima and Anthyllis cytisoides on soil structural stability - A field study in the Rambla Honda Basin, Almería Province, Spain