The effects of forest fire regarding availability of nutrients and iron for plants in a Mediterranean soil

(1)

The effects of forest fire regarding

availability of nutrients and iron for

plants in a Mediterranean soil

Bachelor Thesis Earth Sciences

Name:

Eline Reus

Student number: 10539093

Supervisor:

Dhr. Dr. Julian Campo Velasquez

2

nd

_Supervisor:

_{Dhr. Dr. Boris Jansen}

Date:

June 2016

(2)

2

Samenvatting

De opwarming van de aarde veroorzaakt bosbranden in het oosten van Spanje en bosbranden hebben invloed op de bodemkwaliteit. Erosie verzaakt verlies van nutriënten en zal toenemen doordat er meer zware regenval zal zijn in de toekomst. Bosbranden die zorgen voor verlies van vegetatie zullen ook bijdragen aan toenemende erosie. Nutriënten en ijzer zijn onderzocht in het laboratorium, om zo verschillen te vinden tussen verbrande en niet verbrande bodem. Door het doen van statistische tests, het vergelijken van data en het vergelijken van de resultaten met literatuur is gebleken dat er daadwerkelijk een verschil is. Monsters van een verbrande en een onverbrande heuvel met verschillende eigenschappen en monsters van sedimenten van de verbrande heuvel zijn gebruikt om data te verzamelen met behulp van experimenten. Bosbranden hebben een invloed op bodemkwaliteit afgaande op dit onderzoek. Hoewel de bodemvruchtbaarheid vlak na een bosbrand eerst toeneemt doordat er een vruchtbare laag as ontstaat, wordt dit teniet gedaan door erosie. Erosie gedurende regenbuien na een bosbrand zorgt ervoor dat de bodemkwaliteit achteruit gaat.

Summary

Global warming is related to the increase of forest fires in Eastern Spain, which influence soil quality. Erosion, causing nutrients loss, will be enhanced by heavy rain events in the future and particularly in burned areas (lack of protection by vegetation cover). Nutrients and iron are studied in the laboratory in order to find differences in their availability after a forest fire. To do so, samples with different characteristics from a burnt as well as a control hill are collected and also sediments were sampled in order to study the influence of erosion on nutrient and iron availability. There seems to be a difference between burnt and control soil when doing statistical tests, comparing data and comparing the results with literature. Forest fires decrease the availability of nutrients for plants, because increased soil fertility just after the fire is defeated by erosion during rain events after the fire. Therefore forest fires indirectly lead to a decrease of soil quality.

Abstract

Global warming will cause an increased amount of forest fires in the future in Eastern Spain. This is important to take into consideration, since forest fires do have an influence on soil quality. Forest fires cause an increase in soil nutrient availability, a decrease in iron availability and lead to an erosion increase, because of a lack of protection by a vegetation cover. Nutrient and iron availability are important for plant nutrition and their elimination by erosion will cause decreased soil fertility. Moreover, erosion rates are as well estimated to increase by higher rain intensity which is expected for the near future.

The forest fires induce changes of calcium, potassium, sodium, magnesium, ammonium, nitrate and crystalline and amorphous iron availability in soil. In this study this is determined by measuring soil samples from a burnt hill and a control hill. The data consist of measured salt, acid and water extracts from respectively crystalline iron, amorphous iron and nutrients. The used samples were taken on different hill locations, at different depths and under canopy as well as in bare soil. Erosion is researched with samples from sediments of the burnt hill slope. The location of the study area is nearby Azuébar, a city close to Valencia, in Eastern Spain. Statistical tests, data comparison and comparison with existing literature led to an answer on the research question.

The amount of nutrients and iron available for plants appears to be decreased by forest fires in this Mediterranean area. Forest fire is in this research pointed out as a source of soil degradation. Despite the highly fertile ash layers, that originate from fire and increase soil quality because of their high nutrient content, soil quality is defeated by erosion.

(3)

3

1. Introduction

Global warming has a large indirect influence on soil quality, this is a study that explains part of this statement. To start with: why is there a connection between global warming and soil quality? Global warming means that the average surface temperature will increase and that the average amount of precipitation will change (figure 1.1). Furthermore, there will be heat waves more frequently, which will also last longer (Pachauri et. al., 2014).

Figure 1.1: Change in average surface temperature and precipitation, for two scenario’s: RCP 2.6, which is ... and RCP 8.5, which is... (Pachauri et. al., 2014)

As is shown in figure 1.2, climate change will cause a high probability of forest fires in Europe. Forest fires have a large influence on soil composition. Vegetation damage due to forest fire in summer combined with rain events in autumn, cause a higher amount of erosion of the soil that became bare (Campo et al., 2006; Kosmas et. al., 1997). Since the intensity of precipitation will also increase (Pachauri et. al., 2014), rainfall will cause even more erosion in the near future. Besides this, vegetation is together with climate and topography, the most important factor for soil resilience (Certini, 2005). Moreover, vegetation loss due to forest fires is also appointed as the main cause of soil degradation in Mediterranean areas (Castillo, Martinez-Mena & Albaladejo, 1997). Another influence of forest fire on soil composition seems to be nutrient loss (Certini, 2005), while the ash may cause higher soil fertility (Shakesby, 2011) because of the higher nutrient content. This contradiction is also visible when comparing studies on this issue. Some studies (Badía et. al., 2014) did not found any differences between the amount of nutrients in burnt and unburnt soil. Other studies found differences between soils before and after fire (Knicker et. al., 2005), although it may depend on e.g. the type and intensity of the fire how and how much it changes (González-Pérez et. al., 2004; Ice, Nearly & Adams, 2004). Especially when you take into account these antithetical results, it is of interest to study forest fire influences on soil quality. Soil fertility, or nutrition of vegetation, is important to keep ecosystems healthy, since it is an important ecosystem service (Scröter et. al., 2005). There is a

(5)

5

possibility of change in nutrient and iron availability, caused by forest fire and by the induced erosion after rain events (Kosmas et.al., 1997; Certini, 2005; Ice et. al., 2004).

Figure 1.2: A high confidence of wildfires in Europe, definable out of the figure ‘Widespread impacts attributed to climate change based on the available scientific literature since the AR4’ (Pachauri, et. al., 2014).

One of the studies that investigated this, is the origin of this research and started already in 1995. It consisted of an experiment done in experimental station ‘la Concordia’ in Spain. The aim was to study fire effects on soil quality and erosion of the soil by rain events after the fire (Gimeno‐García, Andreu & Rubio, 2000, 2001, 2004). Therefore samples were taken in different plots, which were burnt at different fire severities. By taking samples and catching erosion sediments after rain events, the influence of forest fire on soil properties was measured. It is proven that vegetation loss causes more erosion during rain events (Campo, 2006,2008).

After that, the researchers decided to start a research on an entirely burnt hill slope to change the scale of the study. Subsequently, a similar hill, which was not burnt, was chosen to be used for gaining control samples, in order to have a reference when determining fire induced changes in soils (Campo, n.d.). The hills are located close to each other in Eastern Spain and are in this research compared in order to find the influence of forest fire on the amount of nutrients and iron in soil. Furthermore, sediment samples of the burnt hill are examined to study erosion induced by rainfall after forest fires. This research contributes to the project mentioned above, since it is a subject that has not been measured yet. Concentrations of amorphous iron, crystalline iron, sulphur, phosphorus, magnesium, calcium, sodium potassium, nitrate, ammonium, phosphate, sulphate, total nitrogen (Ntot), dissolved organic nitrogen (DON) and total organic carbon (TOC)

will be investigated for both hills. In order to explore the distribution of nutrients on both hills separately location, environment and depth are taken into account as well. The aim of this research is to determine the influence of forest fires on soil quality regarding nutrients and iron by comparing the burnt hill with the control hill.

The research question is:

How is the amount of nutrients and iron in soil influenced by forest fires in the Mediterranean area and what does this mean to soil quality?

Sub-questions in order to answer the research question:

- Is there a difference between the amount of nutrients and iron in burnt and in a similar unburnt soil nearby Azuébar in South-Eastern Spain?

- How does a potential difference in the amount of nutrients and iron influence soil quality? - How does erosion play a role in the effect of forest fires on soil composition?

This thesis contains results obtained by using different soil samples for executing experiments for iron and nutrients determination. Which gave insightful information in order to answer the first sub-question by performing statistics. These results are studied with the help of literature in order to determine what the influence of the potential difference on soil quality is. This will lead to an answer to the second sub-question. Furthermore, erosion due to rainfall is taken into account by examining sediment samples that were taken. Comparing this examination with the soil composition on different hill locations, will lead to an answer to

(6)

6

the third sub-question. In the end, there is a synthesis of all the results and discussions in order to answer the research question.

2. Materials and Methods

2.1. Sample Collection

The samples used in this research were collected in Azuébar, a village nearby Valencia in Eastern Spain. They were acquired for the complete study about the influence of forest fires on soil erosion on hill slope scale, which this research is part of. There were samples collected from two hill slopes, one that was recently burnt and one that served as a control hill. Both hills have similar characteristics for soil, vegetation, slope, et cetera.

Figure 2.1: The sampling location in Spain. Figure 2.2: Exact location.

For both hills, the samples were taken on three different locations, namely in the erosion, the transportation and the deposition zone (figure 2.3), so that the effects of processes in the different zones can be studied. For the burnt hill there are extra sediment samples collected after different rain events following the fire: in the autumn of 2014 (November); in spring 2015 (March); in autumn 2015 (Octobre); and lastly in winter 2016 (January). Another difference between the samples is depth on which they are taken, two different depths are used <2 or 2-5 cm. There were no larger depths chosen, since heat is poorly conducted by dry soil and the temperature is not very high in soil at more than 5 centimetres depth (Ice, et. al., 2004). Moreover, there were ‘under canopy’ and ‘bare soil’ samples taken (figure 2.3), since vegetation cover may also have an influence on the soil characteristics.

(7)

7 Figure 2.3: Locations on which the samples were taken on the burnt hill. On the left: the different zones. On the right: under canopy and bare soil samples. The soil movement direction is shown on both sides. The sediments are found logically at the bottom of the hill and were sampled after four different rain events after the fire.

Two samples on two different depths for each hill location were used in this research, which makes twelve samples for both hill slopes, six samples were taken under canopy and six in bare soil. Thus, together with the four sediment samples, 28 samples were studied in total. Since the samples were already used before this research, the samples for both ways of iron determination were already prepared. The samples for nutrient and TOC determination had to be sieved before usage.

2.2. Experiments

In order to collect the data about the soil samples that is needed to answer the research question, three different experiments were executed. The macronutrients could all be measured by making one extract from which electric conductivity (EC) and the total organic carbon content (TOC) were also determined. The EC measurements were helpful for the nutrient measurements described below, since they say something about the values of nutrients to expect. Two different extracts have been made for determining the amount of both kinds of iron. To guarantee more reliability all of the experiments were executed twice for each sample.

2.2.1. Nutrient experiment

To start with, the following nutrients were measured: - Sodium (Na) - Potassium (K) - Magnesium (Mg) - Calcium (Ca) - Phosphorus (P) - Sulphur (S) - Ammonium (NH4) - Nitrate/nitrite (NO2/3)

- Total nitrogen (Ntot)

- Dissolved organic nitrogen (DON) - Sulphate (SO4)

(8)

8

The used experiment was the addition of 50 ml of ELGA-water to 20 g of soil. This mixture was shaken for two hours and after settling for one night, it was shaken 14 hours more. The next day the EC was measured. After filtration (figure 2.4) of the extracts sodium, potassium, magnesium, calcium, phosphorus and sulphur could be measured by inductively coupled plasma mass spectrometry (ICP). The other nutrients were measured with an auto analyzer (AA) and the TOC was measured with a TOC analyzer which is shown in figure 2.5. This resulted in a large amount of data. The data for different locations, environments and depths is suitable to compare to each other and the data from this experiment can be used to compare with the data from the other experiments.

Figure 2.4: Filtration Figure 2.5: TOC-analyzer 2.2.2. Total iron determination

To determine the total content of iron, approximately 1 g of air dry and finely ground soil material was weighed. After weighing 130 ml of sodium citrate was added, together with 2 g of sodium dithionite. This mixture was shaken overnight (figure 2.6) and then 25 ml of a saturated sodium chloride solution was added before diluting it with water to 250 millilitres (figure 2.8). Then the samples were diluted 50 times after settling for 24 hours. The amount of iron was measured by atomic absorption spectroscopy (AAS, figure 2.7). In contrast to the nutrient experiment, this experiment is only done for the control hill and the sediments, since the values for the burnt hill were already measured.

Figure 2.6: Shaking machine

Figure 2.7: AAS Figure 2.8: Solutions in 250ml flasks

2.2.3. Amorphous iron determination

To determine the amount of amorphous iron in the soil samples, 0.125 g of soil sample was weighed (figure 2.9). There was made an extract with 10 ml of an acidified ammonium oxalate solution with a pH of 3.0. The ammonium oxalate reduces the iron and makes it more complex. Since the power of oxalic acid reduces by light the mixture was then shaken packed in aluminium foil, for 4 hours. Then it was centrifuged for half an hour at 3000 rpm. After this the extract was filtrated to make it possible to measure the amount of iron with

(9)

9

ICP. As was the case for the total iron determination, only the control hill and the sediment samples were analysed with this experiments, since the values for the burnt hill were already measured in earlier research.

Figure 2.9: Weighing scale with tubes for the experiment. 2.2.4. Crystalline iron determination

There was no separate experiment in for crystalline iron determination, to get to know the amounts, the amount of amorphous iron has to be subtracted from the total amount of iron. The total amount of iron consists namely out of both crystalline and amorphous iron.

2.3. Data analysis

In order to analyse the data, graphs of the overall results were made in order to compare the concentrations of different nutrients for burnt and control soil and for the different sediments, locations, depths and environments. Furthermore statistical tests were used to compare burnt and control soil and between the different locations, depths and environments for each kind of iron and nutrients. This comparison was done in order to compare the distributions of iron and nutrients on burnt and control hill. Since we use non-parametrical data without a normal distribution, tested with a Kolmogorov-Smirnoff test, a Wilcoxon rank-sum test is used to calculate P-values for each kind of iron and nutrients. Then some trend graphs were made in order to look at differences in trends for the different kinds of iron and nutrients, from which it is possible to predict outcomes of a correlation test. A Spearman correlation test is there used to see whether there is indeed either a positive, negative or weak correlation between both kinds of iron and all of the nutrients. This is a useful tool to find out if iron and nutrients behave the same in control soil as well as in burnt soil.

3. Research Results

The values in figure 3.1 are a summary from the measured values during the experiments for the nutrients concentrations. The same is done for iron in figure 3.2. These figures give a broad overview, since they contain the mean values for nutrients and iron in the samples that are used for this research. Interrelationships between the different values of nutrients and iron are visible. Some nutrient amounts are different in burnt soil, while they are more or less the same in control soil, which mean they behave differently during a forest fire.

(10)

10 Figure 3.1: Nutrients in burnt and control soil.

In the figures 3.1 and 3.2 is visible that there is less iron available in burnt soil, while most nutrients, except NO, are less available in control soil. It is also visible that phosphorus and phosphate occur in very small amounts compared to other nutrients.

3.1. Crystalline and amorphous iron

The figures 3.3 and 3.4 present the mean iron concentrations for control soil. The samples are grouped for the factors depth, environment and location. The means are of four samples, since there are two samples for each combination of factors and every sample is extracted and measured twice.

Figure 3.3: The mean amounts of crystalline iron for samples with different factors in control soil, the blue signs show the standard deviation.

0 50 100 150 200 250 300 350 400 450

S P Mg Ca Na K NO NH4 Ntot DON PO4 SO4 TOC

m

g/

kg

The mean amount of nutrients in burnt and control soil

Burnt Control 0 5000 10000 15000 20000 25000 30000 35000 40000 45000 Amorphous Fe Crystalline Fe m g/ kg

The mean amount of iron in burnt and control soil

0,0 10,0 20,0 30,0 40,0 50,0 60,0 70,0 0-2 0-2 2-5 2-5 0-2 0-2 2-5 2-5 0-2 0-2 2-5 2-5

Down Down Down Down Middle Middle Middle Middle Up Up Up Up

g/

kg

Mean amounts of crystalline iron in control soil

Bare Soil_{Under Canopy}

Figure 3.2: Iron in burnt and control soil. Burnt data from Campo (n.d.)

(11)

11 Figure 3.4: The mean amounts of amorphous iron for samples with different factors in control soil, the blue signs show the standard deviation.

Figure 3.5 and figure 3.6 show the mean iron concentrations for burnt soil. Each bar displays a mean for samples with the same properties for depth, environment and location in control soil.

Figure 3.5: The mean amounts of crystalline iron for samples with different factors in burnt soil (Campo, 2016).

Figure 3.6: The mean amounts of crystalline iron for samples with different factors in burnt soil (Campo, 2016).

In the graphs above it is visible that the values differ more for the crystalline iron. Also crystalline iron occurs in amounts that are around ten times higher than the amounts of amorphous iron.

After analyzing the graphs with mean values a Wilcoxon test for non-parametrical data was used in order to calculate p-values, since the data is not normally distributed and since there are no equal variances

0,0 1,0 2,0 3,0 4,0 5,0 6,0 0-2 0-2 2-5 2-5 0-2 0-2 2-5 2-5 0-2 0-2 2-5 2-5

g/

kg

Mean amount of amorphous iron in control soil

Bare Soil

Under Canopy 0 0,5 1 1,5 2 2,5 3 3,5 4 0-2 0-2 2-5 2-5 0-2 0-2 2-5 2-5 0-2 0-2 2-5 2-5

g/

kg

Mean amounts of amorphous iron in burnt soil

Bare Soil

Under Canopy 0 10 20 30 40 50 0-2 0-2 2-5 2-5 0-2 0-2 2-5 2-5 0-2 0-2 2-5 2-5

g/

kg

Mean amounts of crystalline iron in burnt soil

Bare Soil Under Canopy

(12)

12

between the different groups. Table 3.1 shows the P-values for both kinds of iron and for each factor, respectively environment, depth and different combinations of locations for the control hill.

Control

Crystalline Fe

Amorphous Fe

Environment

0.0166 0.8399

Depth

0.2366 0.0061

Up/Middle

0.5054 0.1949

Middle/Down

0.1304 0.7984

Up/Down

0.0499 0.5054

Table 3.1: P-value table for iron in control soil. The green coloured cells contain P-values that indicate a significant difference.

Table 3.1 shows thus a significant difference for the amount of crystalline iron between the different environments, bare soil and under canopy, and for up/down, which means the erosion zone and the deposition zone regarding location. There is data gained about iron for the burnt soil in earlier research, tests for crystalline iron confirmed significant differences for environment, depth and ‘up/down’ location (Campo, 2016). For amorphous iron the location causes a significant difference when you compare ‘up and down’.

Furthermore, statistical tests are done to look whether there is a significant difference between the values found in burnt soil and the values found. The results of these tests for amorphous and crystalline iron are shown in table 3.2.

Crystalline Fe

Amorphous Fe

Burnt/Control

0.2145 3.66e-05

Table 3.2: P-value table for iron. The green coloured cells contain P-values that indicate a significant difference between the amount of the nutrient in burnt and control soil.

In table 3.2 you can see that there is a significant difference for amorphous iron between burnt and control soil. The amount of amorphous iron is thus significantly higher in control soil.

3.2. Total nutrients

A Wilcoxon test was conducted for the nutrients in order to find significant differences for the locations, environments and depths. Table 3.3 and 3.4, show the P-values for each nutrient and for each factor, environment, depth and different combinations of locations respectively. A significant difference for depth is visible for most nutrients in the table for burnt soil, while a significant difference for environment is visible for most nutrients in the table for control soil. For phosphate in control soil there is a significant difference in depth and for sulphur and nitrate and nitrite in control soil there is a significant difference between up and down slope, or between the erosion and the deposition zone regarding hill location. The difference for the two depths in burnt soil is also visible in figure 3.5 and the difference for the two environments in control soil is shown in figure 3.6.

(13)

13

Burnt S P Mg Ca Na K NO NH4 Ntot DON PO4 SO4

Environment 0.243 0.0649 0.0649 0.1797 0.3939 0.2403 1 0.0931 0.0931 0.1797 0.5909 0.2403

Depth 0.0043 0.0152 0.0152 0.0152 0.0649 0.0152 1 0.0087 0.0087 0.0087 0.0065 0.0043

Up/Middle 1 0.8857 0.8857 0.3429 0.6857 1 1 0.8857 0.8857 0.8857 1 0.8857

Middle/Down 0.4857 0.8857 0.6857 0.3429 0.3429 0.8857 1 0.6857 0.6857 0.4857 0.6571 0.4857

Up/Down 0.4857 1 1 0.4857 0.6857 0.6857 1 1 0.6857 0.4857 0.3143 0.4857

Table 3.3: P-value table for nutrients in burnt soil. The green coloured cells contain p-values that indicate a significant difference.

Control S P Mg Ca Na K NO NH4 Ntot

DON PO

4

SO

4

Environment 0.0411 0.2403 0.0411 0.0411 0.0087 0.0411 0.4848 0.0152 0.1320 0.0931 0.2468 0.0130

Depth 0.3939 0.1320 0.1320 0.1320 0.3939 0.0931 0.4848 0.2338 0.1320 0.0931 0.0476 0.3550

Up/Middle 0.3429 0.2000 0.2000 0.2000 0.4857 0.3429 0.6857 0.3143 0.4857 0.3429 0.3714 0.3143

Middle/Down 0.6857 0.2000 1 0.8857 0.6857 0.8857 0.0571 1 0.3429 0.4857 0.7143 0.8286

Up/Down 0.0286 0.1143 0.2000 0.2000 0.1143 0.2000 0.0286 0.2000 0.0571 0.1143 0.3143 0.1143

Table 3.4: P-value table for nutrients in control soil. The green coloured cells contain p-values that indicate a significant difference.

Figure 3.5: The mean amounts of nutrients in burnt soil at 0-2 and 2-5 cm depth. 0 100 200 300 400 500 600 700 800 900 m g/k g

Mean values for nutrients in burnt soil - different depths

0 - 2 2 - 5

(14)

14 Figure 3.6: The mean amounts of nutrients in bare soil and under canopy, the blue bars show the standard deviation.

Furthermore, a Wilcoxon ranksum test was done to investigate whether there is a significant difference between the amount of nutrients in burnt and in control soil. Table 3.5 shows the P-values that are found for each nutrient.

S P Mg Ca Na K NO

Burnt/Control 0.0017 0.1124 0.0051 0.0304 0.0404 0.0194 1.03e-05 NH4 Ntot DON PO4 SO4 TOC EC

Burnt/Control 0.0032 0.2482 0.0351 0.7468 0.0029 3.66-05 0.0017

Table 3.5: P-value table for nutrients. The green coloured cells contain P-values that indicate a significant difference between the amount of the nutrient in burnt and control soil.

As is visible in table 3.5, the values for sulphur, magnesium, calcium, sodium, potassium, nitrate/nitrite, ammonium, total nitrogen, dissolved organic nitrogen, sulphate, TOC and EC are significantly different for burnt soil compared to control soil. Nitrate/nitrite has significantly higher values in control soil, while the others have significantly higher values in burnt soil.

3.3 Sediments

In figure 3.7 it is visible that the amount of iron increased for each sediment that is produced after the fire. The S1 sample was collected in November 2014, the S2 in March 2015, the S3 in October 2015 and the S4 in January 2016. The first sediment that is collected after the fire has the least iron in it. For the TOC the opposite counts, the last gained sediment contains the least TOC. For the nutrients in figure 3.8 also counts that the first sediment contains the highest amount of nutrients. Furthermore, it is visible that calcium is the nutrient that is found the most in sediments, while phosphorus, ammonium and phosphate occur in the smallest amounts. -200 0 200 400 600 800 1000 m g/ kg

Mean values for nutrients in control soil - different environments

Bare Soil Under Canopy

(15)

15 Figure 3.7: The amount of iron and the TOC in sediments. S1: November 2014, S2: March 2015, S3: October 2015, S4: January 2016

Figure 3.8: The amount of iron and nutrients in sediments. S1: November 2014, S2: March 2015, S3: October 2015, S4: January 2016

3.4. Correlation between iron and nutrients

To see whether behaviour of iron and nutrient availability is similar or contrasting, some trend graphs are made, which are shown in figure 3.9, 3.10 and 3.11. As is apparent from the graphs, the nutrients that are shown, show opposite trends to the iron. This means that iron behaves differently on the different depths, environments, locations and most important on forest fires compared to nutrients. All nutrients that are not displayed have approximately the same trends as those shown, although the available amounts are different for each nutrient.

0 50 100 150 200

Amorphous Fe Crystalline Fe TOC

g/kg

The amount of iron and the total organic carbon in

sediments

S1 S2 S3 S4 0 100 200 300 400 500 600 700 800

S P Mg Ca Na K NH4 Ntot DON PO4 SO4

m

g/k

g

The amount of nutrients in sediments

S1 S2 S3 S4

(16)

16

Environment

Figure 3.9: Trends for different nutrients and iron regarding environment. BS: Bare soil. UC: Under Canopy.

In these trend graphs you see total nitrogen, sodium and sulphur have the same trends. The values are higher under canopy than in bare soil. For the crystalline iron counts that the values in bare soil are higher than the values under canopy. For the amorphous iron the values for burnt soil are higher in bare soil, while the values for control soil are higher under canopy.

Depth

Figure 3.10: Trends for different nutrients and iron regarding depth in centimetres. 0 5 10 15 20 25 BS UC m g/ kg Ntot 0 5 10 15 BS UC Na 0 20 40 60 80 100 BS UC S 36 38 40 42 44 46 BS UC Fe (g/kg) Crystalline Fe 0 1 2 3 4 5 BS UC Amorphous Fe Burned Control 27 28 29 30 31 32 33 0-2 2-5 mg /kg Ntot 0 5 10 15 20 0-2 2-5 Na 0 50 100 150 0-2 2-5 S 32 34 36 38 40 42 0-2 2-5 Fe ( g/ kg) Crystalline Fe 0 1 2 3 4 5 0-2 2-5 Amorphous Fe Burned Control

(17)

17

The trends for depth show that the values are higher in the upper part of the soil for total nitrogen, sodium and sulphur, while the values for crystalline iron are higher at 2-5 cm, this also counts for amorphous iron in control soil. For amorphous iron in burnt soil there is nearly no difference between 0-2 and 2-5 cm.

Location

Figure 3.11: Trends for different nutrients and iron regarding location. ‘Down’ represents the deposition zone, ‘Middle’ the transportation zone and ‘Up’ the erosion zone.

The trends that are based on the different locations show important information about the distribution of iron and nutrients across the slope. The trends are different for burnt soil compared to control soil. Only amorphous iron does not show this difference. Again crystalline iron behaves differently compared to total nitrogen, sodium and sulphur. For the nutrients there are higher values down slope in the deposition zone and lower values in the middle and upper part of the slope which are the transport zone and the erosion zone. For crystalline iron the highest values are found in the middle, while the lowest values are found down slope.

Keeping these trends in mind two correlation images are made and shown in figure 3.12 and figure 3.13. The images show either negative or positive correlations as well as the strength of the correlation between all kinds of nutrients and iron in burnt as well as in control soil. These correlations match with the trends, as was expected.

0 10 20 30 Down Middle Up mg /kg Ntot 0 5 10 15 20 Down Middle Up Na 0 50 100 Down Middle Up S 30 35 40 45 Down Middle Up Fe ( g/ kg) Crystalline Fe 0 2 4 6 Down Middle Up Amorphous Fe Burned Control

(18)

18 Figure 3.12: Correlation image for iron and nutrients in burnt soil. The colours, with corresponding values shown in the bar, show negative or positive correlation.

Figure 3.13: Correlation image for iron and nutrients in control soil. The colours, with corresponding values shown in the bar, show negative or positive correlation.

In the correlation image for burnt soil (figure 3.12), there is a strong negative correlation between crystalline iron and amorphous iron as well as between crystalline iron and all the nutrients. This correlation is less strongly negative for the control soil, and in this soil amorphous iron is also slightly negatively correlating with crystalline iron as well as slightly negatively correlating with all the nutrients. In both the control and the burnt soil all the nutrients are strongly positively correlated. Only NO2/3 is less correlated in the image for

control soil. There is no result for NO2/3, in burnt soil, because the values were too low to measure and

therefore all indicated as lower than 3. This meant the values were all the same in the dataset used, while that is not the case in reality.

(19)

19

4. Discussion

To start with, it has to be mentioned that this research is about two different hills with the same characteristics . One hill is burnt and one hill is used as a control. Parent material, climate, slope and vegetation are similar for the hills and that means possible differences between the burnt and the control soil are forest fire induced.

4.1 The influence of forest fire on nutrients and iron

An interesting result is that there is a significantly different amount of most nutrients after the forest fir. All of them, except NO2/3 are increased after the fire. An increased amount of nutrients in burnt soil is

originated by mineralization of nutrients and the fact that they become a large component of the ash during a forest fire. Although this is not true for nitrogen and phosphorus which e.g. can volatilize at low temperatures (Ice, et. al., 2004; Carter & Foster, 2004). The temperature and the duration of a fire influence the extent of mineralization and volatilisation. Temperatures that are not high enough for volatilization of nutrients cause an immediate increase of their availability. Moreover, nutrients in litter may be ‘locked up’ and they can be released by fire which means they come back in the nutrient circulation (Gray & Dighton, 2006). Plants need nutrition for their metabolism (Barker & Pilbeam, 2015), and an increase in nutrients means thus that the soil becomes more fertile for plants .

A decrease of iron availability in burnt soil could be negatively influencing soil fertility, because iron is an essential micronutrient for plants (Barker & Pilbeam, 2015). Plants need iron for their cell metabolism and iron is very often limiting for their growth, since it is not readily available (Thomine & Lanquar, 2011; Kim & Guerinot, 2007). This means a decrease of iron availability in soils may cause problems for vegetation growth. There is not much known about the behaviour of iron during and after fire (Certini, 2005) and this study can contribute to elimination of this gap, since there is data about the difference in iron content between burnt and control soil available.

4.2 Significant differences compared

Influences of depth and environment are different for burnt soil compared to control soil. In control soil different environments gave significantly different results for nutrients (figure 3.6), the amounts of S, Mg, Ca, Na, K, NH4 and SO4 are significantly higher under canopy. In burnt soil different depths give significantly

different results for nutrients (figure 3.5), the amount of all the nutrients except NO2/3 is significantly higher

at 0-2 cm.

A reason for the fact that environment causes a significant difference in nutrients in control soil, is that vegetation prevents the soil from eroding because vegetation increases the infiltration capacity of the soil (Cammeraat & Imeson, 1999). More important, nutrients in soil are originated from plants and distributed by plants, since litter of plants contains a lot of nutrients and is deposited under the vegetation (Jobbágy &Jackson, 2004). This causes the fact that the amount of nutrients under canopy is significantly higher than in bare soil. Bare soil is more easily eroded and nutrients are vulnerable for erosion (Pimentel et. al., 1995), also there is logically less or no litter deposited on bare soil.

Significant differences in the amount of nutrients for depth in burnt soil also have a likely reason. When there is a forest fire, vegetation is burnt and transformed into ashes, which contains a lot of nutrients as is mentioned earlier. This ash falls on the upper layer of the soil (0-2 cm) and therefore this layer can have a higher nutrient content. As is visible in figure 3.5. This could be a problem, since the upper layer of the soil is the most fertile part of the soil after a forest fire and it is eroded firstly as is discussed later.

In control soil, the amount of nutrients is thus dependent on environment, while depth causes significant differences in burnt soil. This is caused by vegetation creating nutrient availability and fire causing formation of an ash layer on the soil. In control soil there is lack of an ash layer and in burnt soil the vegetation is burnt. Nutrients out of the burnt vegetation spreads more evenly over the whole slope in the form of ash after a fire compared to the environment based spread in control soil.

4.3 NO-availability

There is practically no nitrite (NO2) or nitrate (NO3) found in burnt soil, which is probably because of the fact

that nitrogen is one of the nutrients that is vulnerable to volatilization at relatively low temperatures (Ice, Nearly & Adams, 2004). Other studies found lower NO2/3 values in burnt soil as well (Gimeno-García, et. al.,

(20)

20

2000; Wienhold & Klemmedson, 1992), this is due to leaching, volatilization and water erosion, although the loss depends on the intensity of the fires (Gimeno-García, et. al., 2000).

Nitrification, the process in which ammonium oxidizes and changes into nitrite is a process for which oxygen is normally needed. When nitrite oxidizes it changes into nitrate and that process also needs oxygen. In contrast, the opposite denitrification does not require aerobe circumstances for taking place (Bernhard, 2012). This means that it is not possible that nitrification takes place in anaerobe circumstances during a fire. So, new nitrite and nitrate can not be formed during a fire, which clarifies a part of the results. Since nitrite and nitrate are thus inorganic components of the nitrogen cycle and additionally are one of the primary nutrients needed for life (Bernhard, 2012), it could be interesting to investigate more on this topic.

4.4 The influence of erosion

Nutrients may be removed by wind and water after a fire (Shakesby, 2011), in other words a fire may cause erosion of nutrients. In this research the amount of nutrients in soil is higher after a forest fire than in the control soil (figure 3.1, 3.2), however, it is not certain the nutrients stay in the soil. As was visible in the results (figure 3.7 & 3.8) a lot of nutrients were available in the sediments as well, which means there are nutrients eroded from the hill. Infiltration rates lower due to forest fires and together with loss of vegetation this causes higher erosion rates (Ice, et. al., 2004). Moreover, in eroded soil there is typically a three times larger amount of nutrients available than in the soil on the original place, which means soil fertility is highly impacted by fires (Pimentel et. al., 1995). The nutrient content in burnt soil is therefore higher than in control soil directly after the fire. But since the amount of nutrients in sediments is a lot higher than the amount in burnt soil, it is possible to conclude that the amount of nutrients in burnt soil will decrease soon after the fire.

In other words, fire induces an increased amount of soil nutrients in the form of ash. Unfortunately, this increase can be extinguished by erosion, which is also a result of forest fire. A large part of the nutrients end up in the sediments due to erosion and there is nutrient loss on the slope itself.

4.5 Reasons for correlation

As is already stated, the amount of nutrients in soil increases as a result of forest fire, while the amount of iron decreases. This implies that iron has the opposite behaviour of nutrients when it comes to forest fire. Therefore the results visualized in figure 3.12 and 3.13 are not very surprising. However some interesting observations could be done inquiring the correlation figures.

To start with, the negative correlation of iron with nutrients is less strong in control soil. Therefore, in control soil the nutrients decreases less when the iron value increase than in burnt soil. This gives another difference between burnt and control soil and it therefore gives another indication of an existing difference between burnt and control soil. Secondly, amorphous iron is slightly negatively correlating with the nutrients in control soil, and it also has a slightly negative correlation with nutrients in burnt soil. Lastly the behaviour of nitrite and nitrate is again different from the other nutrients as is visible in figure 3.12. There are no results for this regarding the burnt soil, because the values in the soil were too low to measure and therefore all indicated as lower than three. For the control soil, they are less positively correlated with the other nutrients compared to the average correlation between the other nutrients. This means the behaviour of nitrate and nitrite differs from the behaviour of the other nutrients, this is probably caused by the fast volatilisation of NO2/3 compared to other nutrients (Ice, Nearly & Adams, 2004)

5. Conclusion

5.1 General conclusion

There are significant differences in the availability of iron and nutrients between a burnt and a similar control soil. This difference is positive for nutrients, since they increase which is advantageous for soil fertility. There is no positive influence of forest fire on iron availability at first sight. Erosion by rain events after forest fire causes high nutrient availability in sediments, nutrients are retrieved from the soil on which a highly fertile ash layer existed. This means the fertility of the soil on the slope decreases due to erosion a while after the fire.

(21)

21

Also, the distribution of nutrients changes due to fire. Where the distribution on a control slope is influences by environment, there are found more nutrients under canopy than in bare soil, the depth is more important when it comes to distribution on a burnt slope. It is possible to say the distribution of nutrients becomes vertical due to a forest fire. Regarding the locations of the samples, it is interesting that there are found significantly different amounts of iron down- and upslope, since this also leads to a new distribution.

Concluding, the amount of nutrients and iron is influenced for a large part by erosion after forest fires. And although the fertility of the soil will increase strongly at first, erosion will cause a fast decrease in nutrients availability. Also the fire leads to a new distribution of nutrients and iron, which is not advantageous for the erosion zone of hill and for the soil on 2-5 cm depth. This means soil fertility and therefore soil quality lowers due to forest fire, despite the fact that a lot of nutrients are released by forest fire at first view.

5.2 The state of research

More research regarding the behaviour of iron in soil could be of advantage for the complete study. Especially about iron, there is factually no study regarding forest fires, while iron is a limiting factor in soil fertility.

Moreover, erosion seems to play a major role regarding differences in nutrient and iron availability after a forest fire. Especially when it is compared to the availability before a forest fire. This means it is an interesting subject to investigate more precisely. To do this, taking new samples from the burnt hill a certain period after the forest fire is a solution. With new samples it is possible to investigate how much influence of erosion there has been. This means a new collection of samples that can be analyzed and therefore a lot of new data that can be compared to the already existing data. A large advantage of taking new samples from the same hill is the fact that these new samples can be statistically compared to the current burnt hill samples, since the samples are from the same hill.

6. Evaluation and word of thanks

While executing this research there were a lot of ups and downs. I started the research with enthusiasm, since I found out my research was very important and since I made a nice proposal in which my time schedule showed that there was plenty of time left. I doubted about my English presentation skills, but since I got a seven for my proposal presentation I was content. My proposal is well graded with an eight, while I thought it could be a lot better. This was a sign for me that my perfectionism worked out well and that the doubt I had was not necessary at all. Uncertainty is clearly one of my bad characteristics.

After this first stadium of the research the laboratory work started, which caused a lot more stress. After a week of well going iron experiments I thought that part was finished and on Monday I went to the laboratory to make nutrients extracts. The lab work was well doable for me, because I am very precise, although a little impatient. Unfortunately I had not taken into account that the iron experiments had to be done in duplo and I had to do another week of iron experiments. To make matters worse a shaking machine fell down from the table that week and the next week, I had to start a part of my experiment over again. Luckily during all these failures, I kept quite positive and actually I surprised myself. At that part of the research I got a new motto: ‘When you think you finish tomorrow in the lab, count at least one week extra, than it will only better than expected’. The nutrient extraction went quite smoothly and all the data I needed was there. I was convinced from the fact that I was never going to do lab work again.

Then, after I finished the laboratory work, I started statistics and visualizing data in the week before my presentation. The results were quite nice and my enthusiasm came back, maybe the lab work was good for something in the end. During my presentation a was able to show a lot of tables and figures with clear results and connection between the different groups of data. The final presentation went a lot better than my proposal presentation, which meant I was slightly disappointed with, again, a seven. I learned to accept the fact that I have more writing than presentation skills and full of courage I started writing this thesis. I hope it went out well and I hope that this research really contributes something to the research of Julian Campo, who I will thank for being enthusiastic about my results, for being proud of us after the final presentation, for giving really helpful feedback on my writings and for helping me with all kinds of problems during the research. I will also thank Boris Jansen for giving comments on my proposal and for the feedback after the presentations and I want to thank John Visser for the help in the lab, especially for cleaning the

(22)

22

shaking machine and for keeping it positive. Last but not least, I want to thank my peer students, for their listening ear and for their shared experiences which showed me I was not the only with difficulties.

References

Badía, D., Martí, C., Aguirre, A. J., Aznar, J. M., González-Pérez, J. A., De la Rosa, J. M., ... & Echeverría, T. (2014). Wildfire effects on nutrients and organic carbon of a Rendzic Phaeozem in NE Spain: changes at cm-scale topsoil. Catena, 113, 267-275.

Barker, A. V., & Pilbeam, D. J. (Eds.). (2015). Handbook of plant nutrition. CRC press.

Bernhard, A. (2012). The nitrogen cycle: Processes, players, and human impact. Nature Education Knowledge, 3(10), 25. Cammeraat, L. H., & Imeson, A. C. (1999). The evolution and significance of soil–vegetation patterns following land

abandonment and fire in Spain.Catena, 37(1), 107-127.

Campo, J. (n.d.) Fire in the Mediterranean [PowerPoint slides]. Retrieved from J. Campo. Campo, J. (2016). Data about amorphous and crystalline iron in burnt soil in Azuébar Spain.

Campo, J., Andreu, V., Gimeno-García, E., González, O., & Rubio, J. L. (2006). Occurrence of soil erosion after repeated experimental fires in a Mediterranean environment. Geomorphology, 82(3), 376-387.

Campo, J., Gimeno-García, E., Andreu, V., González-Pelayo, O., & Rubio, J. L. (2008). Aggregation of under canopy and bare soils in a Mediterranean environment affected by different fire intensities. Catena, 74(3), 212-218. Carter, M. C., & Foster, C. D. (2004). Prescribed burning and productivity in southern pine forests: a review. Forest

Ecology and Management, 191(1), 93-109.

Castillo, V. M., Martinez-Mena, M., & Albaladejo, J. (1997). Runoff and soil loss response to vegetation removal in a semiarid environment. Soil Science Society of America Journal, 61(4), 1116-1121.

Certini, G. (2005). Effects of fire on properties of forest soils: a review.Oecologia, 143(1), 1-10.

Gimeno‐García, E., Andreu, V., & Rubio, J. L. (2000). Changes in organic matter, nitrogen, phosphorus and cations in soil as a result of fire and water erosion in a Mediterranean landscape. European Journal of Soil Science,51(2), 201-210.

Gimeno-García, E., Andreu, V., & Rubio, J. L. (2001). Influence of Mediterranean shrub species on soil chemical

properties in typical Mediterranean environment. Communications in soil science and plant analysis, 32(11-12), 1885-1898.

Gimeno-Garcıa, E., Andreu, V., & Rubio, J. L. (2004). Spatial patterns of soil temperatures during experimental fires. Geoderma, 118(1), 17-38.

González-Pérez, J. A., González-Vila, F. J., Almendros, G., & Knicker, H. (2004). The effect of fire on soil organic matter— a review. Environment international, 30(6), 855-870.

Gray, D. M., & Dighton, J. (2006). Mineralization of forest litter nutrients by heat and combustion. Soil Biology and Biochemistry, 38(6), 1469-1477.

Ice, G. G., Neary, D. G., & Adams, P. W. (2004). Effects of wildfire on soils and watershed processes. Journal of Forestry, 102(6), 16-20.

Jobbágy, E. G., & Jackson, R. B. (2004). The uplift of soil nutrients by plants: biogeochemical consequences across scales. Ecology, 85(9), 2380-2389.

Kim, S. A., & Guerinot, M. L. (2007). Mining iron: iron uptake and transport in plants. FEBS letters, 581(12), 2273-2280. Knicker, H., González-Vila, F. J., Polvillo, O., González, J. A., & Almendros, G. (2005). Fire-induced transformation of

C-and N-forms in different organic soil fractions from a Dystric Cambisol under a Mediterranean pine forest (Pinus pinaster). Soil Biology and Biochemistry, 37(4), 701-718.

Kosmas, C., Danalatos, N., Cammeraat, L. H., Chabart, M., Diamantopoulos, J., Farand, R., ... & Mizara, A. (1997). The effect of land use on runoff and soil erosion rates under Mediterranean conditions. Catena, 29(1), 45-59. Pachauri, R. K., Allen, M. R., Barros, V. R., Broome, J., Cramer, W., Christ, R., ... & Dubash, N. K. (2014). Climate Change

2014: Synthesis Report. Contribution of Working Groups I, II and III to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change.

Pimentel, D., Harvey, C., Resosudarmo, P., & Sinclair, K. (1995). Environmental and economic costs of soil erosion and conservation benefits. Science, 267(5201), 1117.

Shakesby, R. A. (2011). Post-wildfire soil erosion in the Mediterranean: review and future research directions. Earth-Science Reviews, 105(3), 71-100.

Stocker, T. F. (Ed.). (2014). Climate change 2013: the physical science basis: Working Group I contribution to the Fifth assessment report of the Intergovernmental Panel on Climate Change. Cambridge University Press.

Thomine, S., & Lanquar, V. (2011). Iron transport and signaling in plants. InTransporters and Pumps in Plant Signaling (pp. 99-131). Springer Berlin Heidelberg.

Wienhold, B. J., & Klemmedson, J. O. (1992). Effect of prescribed fire on nitrogen and phosphorus in Arizona chaparral soil‐plant systems. Arid Land Research and Management, 6(4), 285-296.

(23)

23

Appendix A: Raw data iron

Sample Amorphous Fe Duplo Free Fe Duplo Crystalline

Fe Duplo Environment Location Depth

CA 1-1 4,1 4,3 41,6 36,9 37,5 32,7 UC Up _0-2 CA 1-2 4,6 5,0 13,7 49,4 9,1 44,4 UC Up _2-5,0 CA 2-1 4,6 4,9 51,7 52,8 47,1 47,9 BS Up _0-2 CA 2-2 4,9 5,0 57,0 53,8 52,1 48,8 BS Up _2-5,0 CA 3-1 4,0 4,1 41,6 38,7 37,6 34,6 UC Up _0-2 CA 3-2 5,1 5,6 45,0 36,7 39,9 31,0 UC Up _2-5,0 CA 4-1 4,8 5,2 48,1 31,8 43,3 26,5 BS Up _0-2 CA 4-2 5,1 5,1 54,0 45,3 48,9 40,2 BS Up _2-5,0 CM 1-1 3,8 4,2 28,3 28,2 24,5 24,0 UC Middle _0-2 CM 1-2 4,1 4,5 44,7 43,1 40,5 38,6 UC Middle _2-5,0 CM 2-1 3,9 3,9 37,2 46,3 33,3 42,4 BS Middle _0-2 CM 2-2 4,6 4,4 51,1 47,3 46,5 42,9 BS Middle _2-5,0 CM 3-1 3,8 4,4 38,0 40,9 34,2 36,5 UC Middle _0-2 CM 3-2 4,5 5,1 49,5 47,4 45,0 42,3 UC Middle _2-5,0 CM 4-1 4,5 5,0 51,0 53,7 46,5 48,7 BS Middle _0-2 CM 4-2 5,6 5,5 60,0 63,4 54,4 58,0 BS Middle _2-5,0 CB 1-1 4,7 5,3 53,8 55,6 49,0 50,3 UC Down _0-2 CB 1-2 4,9 5,3 51,9 48,7 47,0 43,3 UC Down _2-5,0 CB 2-1 3,8 4,0 55,8 63,8 52,0 59,9 BS Down _0-2 CB 2-2 4,0 4,2 54,5 61,9 50,6 57,7 BS Down _2-5,0 CB 3-1 3,8 4,5 44,4 45,9 40,6 41,4 UC Down _0-2 CB 3-2 4,5 5,0 46,3 53,9 41,8 49,0 UC Down _2-5,0 CB 4-1 4,1 4,3 41,4 46,9 37,3 42,5 BS Down _0-2 CB 4-2 5,0 5,0 52,3 51,8 47,3 46,7 BS Down _2-5,0 S1 2,6 2,7 23,1 22,7 20,5 20,0 _Sediment S2 3,1 3,0 35,1 38,3 31,9 35,3 _Sediment S3 3,5 3,4 42,6 41,5 39,2 38,1 _Sediment S4 3,5 3,5 45,1 45,6 41,6 42,1 _Sediment

- The values are given in g/kg

- CA/CM/CB #-# samples are control samples. S# samples are sediments samples. #-B are burnt samples. Sample

name with ‘a’ behind it is a duplicate.

- BS: Bare Soil, UC: Under Canopy

- Up: Erosion zone, Middle: Transport zone, Down: Deposition zone

- Depths in centimetres.

Appendix B: Raw data nutrients

Sample

Id

S P Mg Ca Na K NO NH4 Ntot DON PO4 SO4

Bl1 0,52 0,06 0,13 0,69 1,64 0,17 <3 0,27 1,89 1,89 0,02 3,60 Bl2 0,79 0,10 0,06 0,50 1,77 0,09 6,0 0,27 2,24 2,03 0,02 3,60 CA 2-1 11,21 0,55 79,36 284,80 7,71 24,98 123,0 1,40 13,27 7,88 0,07 11,77 CA 2-1a 10,91 0,52 78,10 278,95 8,37 24,46 148,0 1,22 15,80 9,67 0,09 12,97 CA 2-2 9,47 0,28 60,73 182,17 8,84 11,68 104,0 0,72 10,37 6,16 0,05 9,37 CA 2-2a 9,51 0,34 66,35 202,83 9,80 12,89 93,0 0,86 10,68 6,76 0,05 9,13

(24)

24 CA 3-1 65,80 1,57 161,72 737,72 20,04 82,98 900,0 17,19 61,08 16,22 0,28 71,08 CA 3-1a 62,58 1,60 155,14 709,44 18,70 76,36 1110,0 14,21 73,38 23,47 0,31 77,09 CA 3-2 19,21 0,32 60,82 240,00 10,95 32,33 325,0 2,07 20,73 7,74 0,02 15,37 CA 3-2a 20,08 0,37 68,21 269,04 11,74 34,54 248,0 2,21 17,44 7,04 0,05 14,17 CM 2-1 3,10 0,17 25,40 143,34 5,58 5,33 111,0 0,18 8,72 4,83 0,02 1,92 CM 2-1a 3,56 0,25 26,12 148,40 5,56 5,03 92,0 0,18 6,23 3,01 0,05 1,92 CM 2-2 2,10 0,17 15,80 102,45 4,41 1,93 82,0 0,18 6,41 3,54 0,05 1,92 CM 2-2a 3,34 0,25 18,55 116,79 5,04 2,29 67,0 0,18 5,43 3,08 0,02 1,92 CM 3-1 59,00 1,29 141,41 679,19 18,38 73,23 1075,0 12,22 69,03 21,89 0,26 68,20 CM 3-1a 59,93 1,36 147,57 698,72 18,24 77,65 1053,0 13,89 67,53 19,86 0,19 62,92 CM 3-2 9,85 0,24 55,86 168,80 9,54 12,70 132,0 0,99 10,51 5,11 0,02 6,48 CM 3-2a 9,52 0,23 58,06 175,49 10,12 13,70 105,0 1,04 10,82 6,34 0,05 12,97 CB 3-1 7,54 0,41 86,28 312,86 9,34 29,22 48,0 3,11 12,19 8,09 0,09 13,69 CB 3-1a 9,25 0,51 94,72 359,08 10,86 31,85 40,0 3,61 13,10 8,90 0,09 15,37 CB 3-2 3,54 0,07 48,37 164,92 7,58 5,73 44,0 0,50 5,74 3,82 0,05 4,80 CB 3-2a 4,53 0,24 50,81 173,98 7,28 5,84 42,0 0,32 5,64 3,92 0,02 3,60 CB 4-1 2,82 0,19 33,22 149,85 5,33 15,48 38,0 0,36 4,90 3,29 0,05 1,92 CB 4-1a 2,64 0,15 31,30 144,04 5,53 12,39 87,0 0,18 7,08 4,03 0,02 1,92 CB 4-2 3,02 0,18 18,30 76,47 4,43 3,64 85,0 0,18 6,20 3,22 0,02 1,92 CB 4-2a 2,44 0,17 17,98 73,42 4,07 3,33 70,0 0,18 5,46 3,01 0,02 1,92 S1 31,06 4,40 205,57 682,67 19,07 70,81 4,0 5,46 30,82 26,44 4,04 62,20 S1a 31,57 4,03 200,86 653,20 20,22 73,14 4,0 6,00 32,22 27,42 3,89 50,91 S2 18,61 1,25 120,72 448,65 10,50 39,83 <4 3,61 16,64 13,83 1,00 24,26 S2a 20,41 1,25 111,84 421,73 10,23 38,73 10,0 3,25 17,30 14,43 1,07 24,98 S3 10,54 0,81 81,65 400,91 7,46 29,33 <4 3,16 14,50 12,05 0,33 15,85 S3a 11,84 0,82 83,14 411,27 7,25 29,18 <4 3,38 14,64 12,01 0,33 16,33 S4 12,98 0,68 74,66 323,84 9,40 28,30 6,0 2,53 12,36 10,19 0,19 17,29 S4a 12,92 0,68 75,75 326,19 9,09 28,75 4,0 3,16 12,50 9,91 0,21 20,17 1-B 197,85 1,68 279,81 716,79 18,55 146,67 <3 17,05 47,95 34,71 0,36 262,96 1-Ba 192,13 1,54 270,25 686,02 17,84 142,90 <3 18,49 42,14 27,77 0,26 255,76 3-B 25,02 0,33 81,79 206,50 7,80 54,80 <3 4,51 9,77 6,27 0,02 21,61 3-Ba 21,74 0,37 82,74 208,24 7,72 55,65 <3 5,10 10,82 6,86 0,02 16,33 11-B 14,67 0,31 73,20 212,43 9,66 21,90 <3 3,25 7,99 5,46 0,02 10,81 11-Ba 14,57 0,25 70,41 202,53 9,60 21,78 <3 2,89 7,15 4,90 0,02 13,45 15-B 64,37 0,57 108,83 415,94 17,74 59,96 <3 9,70 18,91 11,38 0,05 75,65 15-Ba 65,30 0,56 106,27 380,98 17,47 58,69 <3 10,19 19,26 11,35 0,05 73,97 17-B 18,79 0,34 109,50 297,99 9,57 21,17 <3 4,74 10,16 6,48 0,02 19,93 17-Ba 18,46 0,32 114,72 308,11 9,47 22,18 <3 5,32 10,93 6,79 0,02 14,41 19-B 16,22 0,37 64,77 209,10 14,23 34,15 <3 3,47 6,86 4,17 0,05 11,77 19-Ba 14,75 0,24 64,36 219,94 13,44 32,21 <3 3,61 7,67 4,87 0,02 8,17 28-B 169,92 1,03 177,83 747,64 26,11 91,22 <3 18,94 39,51 24,80 0,17 219,50 28-Ba 190,21 1,17 198,92 841,70 27,89 96,70 <3 19,44 44,97 29,88 0,19 266,33 29-B 118,44 0,95 180,38 558,08 15,07 73,85 <3 8,93 22,70 15,76 0,07 110,71 29-Ba 122,17 1,08 180,70 574,07 16,25 77,51 <3 6,59 32,01 26,90 0,14 175,31 30-B 28,98 0,25 81,70 265,64 8,76 19,68 <3 2,26 8,16 6,41 0,05 33,86 30-Ba 29,74 0,26 80,04 253,65 8,32 19,87 <3 2,03 9,11 7,53 0,05 35,54

(25)

25 31-B 145,68 1,10 247,53 671,90 16,03 65,24 <3 17,09 44,45 31,17 0,14 184,20 31-Ba 176,86 1,32 263,86 731,53 17,35 63,55 <3 17,32 48,37 34,92 0,19 240,63 36-B 35,35 0,31 87,24 258,22 10,31 29,22 <3 5,64 13,94 9,56 0,07 37,22 36-Ba 34,64 0,32 89,63 262,73 10,26 30,41 <3 5,50 14,26 9,98 0,05 37,94 40-B 50,31 0,37 103,16 395,41 18,98 53,29 <3 8,34 21,05 14,57 0,05 62,20 40-Ba 52,14 0,49 110,90 414,25 20,04 58,55 <3 8,03 21,79 15,55 0,05 62,44

- Data in mg/kg, except NO (NO2 + NO3) is in µmol/l.

- Up: Erosion zone, Middle: Transport zone, Down: Deposition zone

Appendix C: Raw data TOC/EC

Sample code EC (µS/L) EC2 TOC (g/kg)

TOC2 Environment Location Depth CA 2-1 410 412 70,98 68,52 BS Up 0-2 CA 2-2 320 349 27,44 27,44 BS UP 2-5,0 CA 3-1 823 791 549,1 553,9 UC Up 0-2 CA 3-2 361 376 74,79 76,5 UC UP 2-5,0 CM 2-1 226 218 8,683 7,717 BS Middle 0-2 CM 2-2 195 201 3,972 4,198 BS Middle 2-5,0 CM 3-1 797 795 424,5 445,3 UC Middle 0-2 CM 3-2 290 299 22,65 25,29 UC Middle 2-5,0 CB 3-1 372 401 79,28 82,78 UC Down 0-2 CB 3-2 231 236 20,97 20,2 UC Down 2-5,0 CB 4-1 245 237 11,14 10,33 BS Down 0-2 CB 4-2 158,5 156,6 5,06 3,939 BS Down 2-5,0 S1 777 656 S2 466 453 167,2 145,6 S3 461 445 136,2 136,4 S4 441 481 103,8 108,3 31-B 958 921 649,9 766,7 UC Up 0-2 17-B 533 612 128,2 135 UC Up 2-5,0 15-B 724 739 162,8 160,3 BS Up 0-2 19-B 439 456 47,1 52,39 BS Up 2-5,0 1-B 1378 974 721,9 639,3 UC Middle 0-2 3-B 485 470 79,65 96,27 UC Middle 2-5,0 36-B 565 548 87,73 93,71 BS Middle 0-2 11-B 406 406 52,79 48,06 BS Middle 2-5,0 28-B 1063 1016 517 597,7 UC Down 0-2 40-B 737 668 141,6 152,2 UC Down 2-5,0 29-B 996 694 345 331,8 BS Down 0-2 30-B 484 468 66,39 70,58 BS Down 2-5,0

(26)

26

Appendix D: Matlab scripts

Appendix D.1 : Statistics script nutrients

%%%%% Eline Reus

%%%%% Bachelor Thesis Aardwetenschappen %%%%% Juni 2016

%%%%% Is the data for the different locations, depths and environments %%%%% significantly different? Kruskal-Wallis tests and ranksum tests. %%%%% In this script Sulfur can be exchanged for every other nutrient.

%% %%%%% Initialisation clc clear close all addpath('C:\Users\Eline\Documents\3FPS') load 'OverviewBurned.mat' load 'OverviewControlS2.mat' %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% %% Sulfur %% Parametric/Nonparametric data? % Normal distribution? MeanS = OverviewBurned.MeanS

KSTEST = kstest(MeanS); % Answer = 1, not normal distributed

%% Locations

Up = strcmp(OverviewBurned.Location,'Up');

UpSamples = OverviewBurned((Up==1),:) ;

Middle = strcmp(OverviewBurned.Location,'Middle');

MiddleSamples = OverviewBurned((Middle==1),:);

Down = strcmp(OverviewBurned.Location,'Down');

DownSamples = OverviewBurned((Down==1),:) ;

Sediment = strcmp(OverviewBurned.Location,'Sediment');

SedimentSamples = OverviewBurned((Sediment==1),:);

% Create histograms to look at the distribution.

figure(1) subplot(1,3,1) hist(UpSamples.MeanS) subplot(1,3,2) hist(MiddleSamples.MeanS) subplot(1,3,3) hist(DownSamples.MeanS)

% Not normal distributed data. Use non-parametrical statistical tests.

% Test if Up/Middle location values differ significantly

TestUMMeanS = ranksum(UpSamples.MeanS, MiddleSamples.MeanS);

% Answer = 1, so there is no significant difference between the % distribution of data between the three locations.

% Test if Middle/Down location values differ significantly

TestMDMeanS = ranksum(MiddleSamples.MeanS, DownSamples.MeanS);

% Answer = 0.4857, so there is no significant difference between the % distribution of data between the three locations.

% Test if Up/Down location values differ significantly

(27)

27

%% Environments UC = strcmp(OverviewBurned.Environment, 'UC'); UCSamples = OverviewBurned((UC==1),:) ; BS = strcmp(OverviewBurned.Environment, 'BS') ; BSSamples = OverviewBurned((BS==1),:);

figure(2) subplot(1,2,1)

hist(UCSamples.MeanS) subplot(1,2,2)

hist(BSSamples.MeanS)

% Test if under canopy/bare soil environment values differ significantly

TestEnvironmentMeanS = ranksum(UCSamples.MeanS, BSSamples.MeanS)

%% Dephts LowDepth = strcmp(OverviewBurned.Depth, '0-2'); LowDepthSamples = OverviewBurned((LowDepth==1),:); HigherDepth = strcmp(OverviewBurned.Depth, '2-5,0'); HigherDepthSamples = OverviewBurned((HigherDepth==1),:);

figure(2) subplot(1,2,1)

hist(LowDepthSamples.MeanS) subplot(1,2,2)

hist(HigherDepthSamples.MeanS)

% Test if low and higher depth values differ significantly

TestDepthMeanS = ranksum(LowDepthSamples.MeanS, HigherDepthSamples.MeanS)

Tests = {'Up/Middle', 'Middle/Down', 'Up/Down', 'UC/BS', 'Depth', 'KSTest'}

Results = [TestUMMeanS, TestMDMeanS, TestUDMeanS, TestEnvironmentMeanS,

TestDepthMeanS, KSTEST]

ResultsMeanS = table(Results)

Appendix D.2: Statistics script iron

%%%%% Eline Reus

%%%%% Bachelor Thesis Aardwetenschappen %%%%% Juni 2016

%%%%% Is the data for the different locations, depths and environments

%%%%% significantly different for iron? Kruskal-Wallis tests and ranksum tests.

%% %%%%% Initialisation clc clear close all addpath('C:\Users\Eline\Documents\3FPS')

(28)

28 load 'OverviewIron.mat' %% Parametric/Nonparametric data? % Normal distribution? AmorFe = [OverviewIron.AmorFe,OverviewIron.AmorFe2]

MeanAmorFe = mean(AmorFe,2) % Calculate mean from first

and duplo sample

kstest(MeanAmorFe); % Answer = 1, not normal

distributed

CrysFe = [OverviewIron.CrysFe,OverviewIron.CrysFe2]

MeanCrysFe = mean(CrysFe,2) % Calculate mean from first

and duplo sample

kstest(MeanCrysFe); % Answer = 1, not normal

distributed % Homogenous variance? VarMeanAmorFe = vartestn(MeanAmorFe); VarMeanCrysFe = (std(MeanCrysFe))^2; % Data is non-parametric %% Amorphous Iron

%% Split Samples on different properties to test whether these properties cause significantly different values.

%% Locations

Up = strcmp(OverviewIron.Location,'Up');

UpSamples = OverviewIron((Up==1),:) ;

Middle = strcmp(OverviewIron.Location,'Middle');

MiddleSamples = OverviewIron((Middle==1),:);

Down = strcmp(OverviewIron.Location,'Down');

DownSamples = OverviewIron((Down==1),:) ;

Sediment = strcmp(OverviewIron.Location,'Sediment');

SedimentSamples = OverviewIron((Sediment==1),:);

% Make one column for the mean of the first measurement and the duplo

MeanUpAmorFe = mean([UpSamples.AmorFe,UpSamples.AmorFe2],2);

MeanMiddleAmorFe = mean([MiddleSamples.AmorFe,MiddleSamples.AmorFe2],2); MeanDownAmorFe = mean([DownSamples.AmorFe,DownSamples.AmorFe2],2);

%Test whether there is equal variation

VarLocationSamples = vartestn([MeanUpAmorFe,MeanMiddleAmorFe,MeanDownAmorFe]) ;

% Answer: p-value = 0.89914

%Test whether there is normal distribution

DistrLocationSamples = kstest([MeanUpAmorFe,MeanMiddleAmorFe,MeanDownAmorFe]) ;

% Answer = 1, not normal distributed

figure(1); subplot(1,3,1) hist(MeanUpAmorFe) subplot(1,3,2) hist(MeanMiddleAmorFe) subplot(1,3,3) hist(MeanDownAmorFe)

The effects of forest fire regarding availability of nutrients and iron for plants in a Mediterranean soil