The long-term fate of deposited 15N in a coniferous forest in Alptal, Switzerland

(1)

BSc Thesis

The long-term fate of deposited

15

_{N in a coniferous}

forest in Alptal, Switzerland

(2)

BSc Thesis – The long-term fate of deposited 15N in a coniferous forest in Alptal, Switzerland Quirine Cohen – 10516794 – 05 May 2018 – University of Amsterdam

Abstract

The deposition of nitrogen (N) in forest ecosystems has increased due to anthropogenic activities such as industrial combustion and agriculture. These increasing depositions of inorganic N can lead to various effects on the soil and vegetation such as acidification, enhanced tree growth or suppressed decomposition of soil organic matter. Moreover, N may influence the carbon (C) cycle of an ecosystem, due to its strong link to C. This dependency is determined by the long-term fate of N. C sequestration as a result of N deposition can be quantified by using 15

N as a tracer in field experiments. In order to investigate the impact of N deposition on forest ecosystems, N saturation field experiments were set up in parts of West-Europe as an integrated research project in the early 1990s (NITREX). Tietema et al. (unpublished results) investigated the long-term effects of N deposition by investigating N retention of labelled NH4

+

after 20 years. This research recovered most of the N in the mineral soil layer in contrary to the previous studies (Koopmans et al. 1996; 1997, Wessel et al 2013). A similar experiment was done in the Alptal region in Switzerland in 1994 with 15

NH4NO3. Measurements were done on 2 catchments with different N additions and altering slope gradients. In 1995 a second research was done, measuring N retention after adding NH4

+

and NO3 -

separately. This research investigates the 15

N retention, 23 years after the first N addition. Samples from both catchments were tested on moisture content, pH, electrical conductivity (EC) and slope gradient. Moreover, 15_{N retention and C and N content were} measured using an Elemental Analyser coupled to an Isotope ratio mass spectrometer. The major N sink in these areas were the Lfh horizons and decreased from the A to B-horizon. The Lfh horizon remained the major 15

N sink when comparing with A and B separately. When combining the A and B horizon, the major 15

N sink was the mineral soil, but the difference was insignificant. The C:N ratio was significantly different in the Lfh horizons between the two catchments. Moreover, correlations were found between EC and 15

N in both catchments, as well as strong correlations between EC and C:N ratio.

(3)

Table of content

INTRODUCTION ... 4 METHODS ... 5 Study Area ... 5 Laboratory analyses ... 6 Calculations ... 6 Statistics ... 7 RESULTS ... 8 Major 15_{N sink ... 8}

The effect of continuous N addition ... 10

Soil moisture content, pH, EC and slope gradient ... 11

DISCUSSION ... 16 CONCLUSION ... 18 REFERENCES ... 19 APPENDIX ... 21 A.1. P –VALUES ... 21 A.2. CORRELATION GRAPHS ... 22 A.3. MATLAB SCRIPTS ... 23

(4)

Introduction

Due to anthropogenic activities such as industrial and automotive combustion or intensive agriculture, the amount of deposited N in forest ecosystems has increased (Templer et al., 2012, Dise & Wright, 1995). Depositions of inorganic N can have various effects on these forest ecosystems such as enhanced tree growth, suppressed decomposition of soil organic matter and acidification of soils and streams (Goodale 2016). Acidification can increase nitrate (NO3

-) in runoff either directly through N input or indirectly by root uptake. This is a result of increased anion movement accompanied by loss of nutrient cations. When the soil is N saturated and nitrification takes place, the nitrate concentration will increase and total anion leaching will be lost. In the Great Smoky mountains where spruce is ambient, nitrate leaching has increased after elevated losses of aluminium (Aber et al., 1989). Since N is linked to C, an increase in N deposition can influence the C cycle in a certain system. C sequestration as a result of N deposition can be estimated using 15

N tracers (Tietema, unpublished results).

In order to study N cycling depending on atmospheric input of N, the NITREX project was set up under the Commission of the European Communities (CEC) program of Science and Technology for Environmental Protection (STEP) (Dise and Wright 1992, 1995). In total 9 sites were investigated with either removal or addition of N (Dise & Wright, 1992). Most studies focused on the 15

N retention after 1-2 years and recovered the majority of 15 N in litter and the organic soil. Other studies that investigated 15

N retention on unfertilized catchments after 7 to 9 years also recovered most of the labelled N in litter and in the organic soil (Goodale 2016).

One has investigated the long-term fate on N deposition 1, 9 and 20 years, after addition of labelled NH4+

, which was part of the NITREX project (Koopmans et al., 1995, Wessel et al. 2013, Tietema, unpublished results). These researches were conducted in a forest near Ysselsteyn, the Netherlands with an acidic and well-drained haplic podzol soil. After one year, most of the 15

N (60%) was recovered in the organic soil layer and decreased with depth (C. J. Koopmans, Lubrecht, and Tietema 1995; Wessel, Tietema, and Boxman 2013). In the mineral soil layer (upper 50 cm) 20% of the labelled N was retained. After 9 years, in 2001, these measurements were done again. The 15

N recovered in the mineral soil layer was only 2% and in the organic soil layer, about 40% was recovered. In 2012 the amount of recovered 15

N in the mineral soil layer increased to 28%, about 6% more than in the organic soil layer (Tietema et al., unpublished results). This was in contrast with previous years, when most of the 15_{N was retained in the organic soil layer (Koopmans et al., 1996; Wessel et al., 2013).}

In order to investigate if the long-term fate of 15

N retention in the mineral soil is also valid for other locations, a catchment in Alptal, Switzerland was investigated. Two other studies focused on the Alptal region 1 year and 3 years after 15

N addition (Schleppi et al., 1998, Providoli et al., 2005). In the research conducted by Schleppi et al. (1999) 3 years after first 15

N addition, which was also part of the European research project NITREX, the 15

N retention was higher in the organic layer than in the mineral soil layer. In the second research, conducted by Providoli et al., (2005) one year after first 15

N addition, most of the 15

N was recovered in above ground vegetation. For the NH4 +

tracer, 10.2% was recovered in the Lfh horizon and 4.3% in the A horizon, no 15

N was recovered in the B horizon. For the NO3

-

tracer 11.2% was found in the LF horizon, 2.8% in the A horizon and 16.8% in the B horizon. Nonetheless, this increase since 1994 in 15

N retention was expected to be related to a dry winter in 2001 when cracks were formed in the clay. During snowmelt, 15

N leached into the B horizon and remained there after the closing of these cracks (Providoli et al. 2005).

(5)

The objective of this research was to investigate the effects of 15

N deposition on long-term 15_{N retention in soil profiles of two catchments in the Alptal valley and in what way different} N addition levels and other regulatory factors could influence this. In order to investigate this, the following research questions were proposed.

1. Which soil layer (Lfh, A or B horizon) is the major 15

N sink, 13 (catchment 1) and 19 years (catchment 2) after 15

N deposition?

2. What is the effect of continuous increased N addition on 15

N retention in the soil? 3. How is 15

N retention influenced by soil moisture content, pH, EC and slope gradient?

Methods

Study Area

The research was carried out in the Alptal region in Switzerland, which lies on the northern side of the Alps at 1200m height with a cool and wet climate. The mean annual precipitation is 2300 mm and the mean annual temperature is 6˚C. The vegetation in the area consists mainly of the 250-year-old species, Norway spruce (Picea Abies), and was never artificially fertilized. Other plant species present are silver fir (Abies Alba) and Vaccinium myrtillus. Ground vegetation consists of Poa trivialis, Petasites albus. Carex ferruginea and Caltha palustris. The atmospheric deposition of inorganic N is equally distributed between NO3

-and NH4

+

, to a total of 12 kg N ha-1 yr-1

. The major soil types in the area are umbric Gleysols and the parent rock is Flysh, which are sedimentary conglomerates with clay-rich shists (Schleppi et al., 1999). The slope has a west aspect and is about 20%, however there is a slight difference between the slopes in both catchments. There are 3 horizons; a Lfh, A and B horizon which have a very low permeability (Providoli et al., 2005).

The two catchment areas that were used in this research are approximately 1600m2 and delimited by trenches. The water balance in these catchments is almost balanced since the soil has a low permeability and therefore the water will not infiltrate below these trenches, to about 80cm. Catchment 1 received only ambient N deposition (12 kg N ha-1

yr-1 ), while catchment 2 was fertilized with NH4NO3 from 1995 until present (fig. 1).

The N was added using forty rotating sprinklers each 1.5 meters above the ground. In order to simulate a deposition increase of atmospheric N, small additions were made during the simulated precipitation events. These events were simulated by collecting water on a polyethylene sheet of 300 m2 in the forest (Schleppi et al., 1999). Catchment 2 was labelled with 15_NH

4NO3 during one year from 1995-1996 (219 mmol m

-2_{0.88 atm%}15_{N). Catchment} 1, which was initially the control catchment without label, was labelled with K15

NO3 (0.17 mmol m-2 99 atm%15 N) in 2000-2001 and with 15 NH4CL in 2002-2003 (0.7 mmol m -2 99 atm%15

N). The applied tracers were low to minimize the fertilization effect compared to ambient N deposition. To mimic seasonality, the sprinklers were replaced with a backpack-sprayer during winter and applied occasional concentrated NH4NO3 solution.

From catchment 1, 24 cores and from catchment 2, 20 cores were taken, using cylinders with a 5 cm diameter. Each core was separated in 3 horizons (Lfh, A and B) resulting in 72 and 60 samples per catchment.

(6)

Figure 1. Catchment 1 and 2 in Alptal, Switzerland (Providoli et al. 2005).

Laboratory analyses

The samples have been sieved with a 2 mm sieve and have been dried in an oven at 40 °C. Afterwards, each mineral sample was moulded with a ball mill and each organic sample was moulded with a centrifuge mill. Lfh, A and B horizon samples were then weighed (10 mg, 20-25 mg, 30-60 mg respectively) and folded in threefold in tin capsules. With these capsules total N and C were measured using an Elemental Analyser (EA), which is coupled to an Isotope ratio mass spectrometer (IRMS) to measure 15

N. Acetanilide was used as a standard. Reference gas (high-purity N2 gas) was calibrated to atmospheric N2 standard (at-air) using certified reference materials (IAEA-N2 and IAEA-NO3; from the Department of Nuclear Applications International Atomic Energy Agency, Vienna). In order to measure pH and EC a water extract was made. Samples were already dried at 70 °C and 105 °C to calculate soil moisture content of the organic and mineral soil respectively.

Calculations

The abundance of labelled N in an N pool can be indicated in two manners. The first is as delta 15

N (δ15

N), the second as a percentage of the total N added. Delta 15

N is determined as: δ15 N = (_!!!!"#$% !"#$%#&% - 1 ) x 1000‰ R = ratio of 15 N to 14

N of the sample. The R value of atmospheric N2 is used, which is 0.0036764. The percentage of labelled N can be determined by:

at.% 15

N = _{!! !}!!"#$%&

!"#$%& x 100%

The total N deposition in a pool can be calculated as a percentage with the following equation derived from Wessel et al. (2013):

%15 Nrec,i=

!!!".%!"!!!!".%!"!!,!"!#

!!"#$!

%15

N is the mass atom percentage of 15

N tracer recovered in the soil pool i. With at.% 15

N as

15

N abundance in N pool I and at% 15

Ni,init as a reference in soil pool i. before 15

N was added. Pool sizes (mi) were used as amount of N mass (kg N ha

-1

(7)

The added 15

N (mlabel)was the total amount (kg 15

N ha-1

) of the added tracer (Wessel, Tietema, and Boxman 2013).

Statistics

In order to determine significant differences between the horizons of the two catchments, a two-way ANOVA was conducted using Matlab. This test was done for 15

N recovery and C:N ratio’s. A significance level of α < 0.05 was used for each variable. The following hypotheses were determined:

H0 : The means are not significantly different Ha: The means are significantly different. To determine the linear correlation between 15

N recovery and pH, EC, moisture content, C:N ratio and EC in both catchments, scatterplots were made. Moreover, the Pearson's product moment correlation coefficients (r) and coefficients of determination (r2

) were calculated. The strength of the correlation becomes stronger when r gets closer to -1 or 1 (Burt, Barber, and Rigby 2009).

(8)

Results

Major 15_{N sink}

The total recovery of 15

N was highest in catchment 1 (98.7%) in comparison with catchment 2 (47.3%). In catchment 1, 49% of the 129.2 g/ha added 15

N was recovered in the organic soil layer and less 15

N was recovered in the A and B horizon, 31.6% and 18.2% respectively (fig 2). The differences between the horizons in catchment 1 were insignificant (Appendix A.1). In comparison with the research of Providoli et al. (2005), the 15

N recovery in catchment 1 has increased with overall 53.4% within 19 years. For each catchment the average C:N ratio was determined per horizon. Overall, the first catchment had the highest C:N ratio’s, decreasing per horizon (table 1). The mean C:N ratio’s of NO3

-

and NH4 + combined in catchment 1 per horizon were 27,8, 18,1 and 15,2 from the Lfh to A and B horizon. In 2002/2003 the C:N ratios were 21, 17.1 and 17.7 respectively (fig. 3). The difference in C:N ratio’s in 2016 between the organic soil horizon and A horizon and the difference between the A and B horizon were both significant (Appendix A.1). The recovery of 15_{N in catchment 2 was lower for each horizon than in catchment 1. In catchment 2,} 20.2% 15

N was recovered in the Lfh horizon, 18.3% in the A horizon and 8.8% in the B horizon (fig 4). The difference between the organic soil horizon and the B horizon was significant (p < 0.008) in catchment 2. The differences between the other soil layers were all insignificant (Appendix A.1). In comparison with the 15

N retention in 1997, the total 15 N recovery in 2016 decreased with 28.7% (fig. 4). This decrease in 15_{N retention apropos of} 1997 can also be seen in the C:N ratio, which declined to 17.9 in 2016 (fig. 5). This C:N ratio is slightly higher, but comparable with the C:N ratio in 2002. The 15

N recovery in the organic layer at catchment 2 was higher in 2016 in comparison with 1997, but decreased with 36% in the A and B horizons combined (fig 5). In both catchments the major 15

N sink was the organic soil layer and the highest overall 15

N recovery was in catchment 1.

Figure 2. %15_{N recovery in 2002/03 & 2016 (Providoli et al. 2005).}

21,4 7,1 16,8 49,0 31,6 18,2 0 10 20 30 40 50 60 70

Mean Lfh horizon Mean A horizon Mean B horizon

% 15N re cove ry Catchment 1 - 15_{N recovery (%)} 2002/03 & 2016 2002/2003 2016

(9)

Figure 3. C:N ratio’s in 2002/03 & 2016 (Providoli et al. 2005).

Figure 4. %15_{N recovery in 1997 and 2016 (Schleppi et al. 2004).}

Figure 5. C:N ratio’s 1997, 2002 & 2016 (Schleppi et al. 2004).

21,0 17,1 17,7 27,8 18,1 _15,2 0,0 5,0 10,0 15,0 20,0 25,0 30,0 35,0

Mean Lfh Mean A Mean B

C: N ra ti o Catchment 1 - C:N ratio's 2002/'03 & 2016 2002/'03 2016 13 42 21 20,2 _18,3 8,8 0 10 20 30 40 50 60

Mean Lfh horizon Mean A horizon Mean B horizon

% 15N re cove ry Catchment 2 - 15_{N recovery (%)} 1997 & 2016 1997 2016 20,2 17,18 17,9 15 16 17 18 19 20 21

Mean total catchment

C: N ra ti o Catchment 2 - C:N ratio's 1997 / 2002 / 2016 1997 2002 2016

(10)

Long-term N treatment effects on 15

N retention in the soil.

The long-term effects of increased N deposition on 15_{N retention in the soil were} investigated with 19 years of increased N deposition in catchment 2 while catchment 1 received ambient N deposition. In figure 6 the differences in 15

N recovery between the two catchments are shown. The left beams show the results for catchment 1 and the right beams catchment 2. The differences are visualized for the Lfh horizon, A horizon, B horizon and the total soil layer. In all soil horizons 15

N recovery was higher in catchment 1 in comparison with catchment 2. This is remarkable since catchment 2 has received 19 years of increased N deposition and is probably due to N leaching. The total 15

N recovery was significantly (p < 0.0101) higher in catchment 1 in comparison with catchment 2. In the Lfh horizon these differences were also significant (p < 0.0236), but in the A and B horizon these differences were insignificant. When looking at the differences in C:N ratio’s between the catchments (fig 7), again the organic layer showed significant differences (p < 0.000002) between the catchments. The C:N ratio in the total soil was lower in catchment 2, however insignificant (p < 0.1042).

Figure 6. % 15_{N recovery in catchment 1 & 2, 2016. *p level difference Lfh horizons = 0.0738. **p level}

difference A horizons = 0.5767. ***p level difference B horizons = 0.9292.

49 31,6 18,2 98,7 20,2 _18,3 8,8 47,3 0 20 40 60 80 100 120 140

Llf Horizon A horizon B horizon Total %15N recovery

% 15N

Catchment 1 & 2 - 15_{N recovery (%)}

2016 Catchment 1 (Low N addition) Catchment 2 (High N addition)

(11)

Figure 7. C:N ratio’s in catchment 1 & 2, 2016. *p level difference Lfh horizons = 0.0000. **p level difference A horizons = 0.993. ***p level difference B horizons = 0.9981.

Soil moisture content, pH, EC and slope gradient

Using scatterplots soil moisture content, pH, EC and slope gradient were measured to investigate possible correlations with 15

N retention in the soil. These factors can influence the degree of 15

N retention in the soil and are shown in figure 8, 9, 10 and 12. In figure 13 the mean %15

N retentions per hill slope are displayed, the left beam shows catchment 1 and the right beam shows catchment 2. Table 2 shows the mean and standard error of the mean (SEM) values of soil moisture content, pH and EC, per soil horizon, and for each catchment separately. See Appendix A.1 for p-values of every variable. In the first catchment the moisture content decreased significantly from the organic soil horizon to the B horizon and from the A to B horizon. The pH decreased from the Lfh horizon to the A horizon and increased from the A to the B horizon, all differences were insignificant. The electrical conductivity almost halved, significantly, from the Lfh to A horizon and remained fairly the same, insignificantly, from the A to B horizon, 148.9 and 148.4 respectively. In catchment 2 the moisture content decreased per horizon but was only significant from the Lfh to B and A to B horizon. The pH increased per horizon and was significant in the same horizons as the moisture content. Also, the electrical conductivity decreased with depth per horizon and was significant except from the A to B horizon.

Scatterplots were made to investigate the correlations between 15

N recovery and the soil variables. In catchment 1, a weak, to moderate positive correlation was found between the electrical conductivity and 15

N recovery. The Pearson’s product-moment correlation coefficient was 0.47 and the determination coefficient was 0.22. This implies that the EC ‘explains’ 22% of the 15

N recovery. Moreover, a strong correlation was found between C:N ratio and EC, with Pearson's r of 0.58, and a determination coefficient of 0.34. No correlation was found between pH and 15

N and moisture content and 15

N. In catchment 2, a weak correlation was found between pH and 15

N. The Pearson’s r was -0,20 and the determination coefficient was 0,038. Therefore it can be stated that pH ‘explains’ 3,8% of the 15

N recovery. Another moderate positive correlation was determined between EC and 15

N. The Pearson’s r was 0,43 and the determination coefficient 0,19. No correlation was found between moisture content and 15

N recovery. In the correlation calculation between EC and C:N ratio, a Pearson’s r was found of 0,44 and a determination coefficient of 0,20. This means there is a moderate positive correlation between EC and C:N ratio.

27,8 18,1 _15,2 19,7 21,7 17,7 _15,9 17,9 0,0 5,0 10,0 15,0 20,0 25,0 30,0 35,0

LOh horizon A horizon B horizon Mean total catchment

C:

N

ra

ti

o

Catchment 1 & 2 - C:N ratio's 2016 Catchment 1 (Low N addition) Catchment 2 (High N addition)

(12)

Figure 8. Correlation between EC and 15_{N in catchment 1, 2016. r}2 _{= 0,22 and r = -0,47.}

Figure 9. Correlation between C:N ratio and EC catchment 1, 2016. r2_{= 0,58 and r = 0,76.}

Figure 10. Correlation between pH and 15_{N in catchment 2, 2016. r}2_{= 0,46 and r = -0,68.}

y = -0,892x + 47,53 r² = 0,21916 0 20 40 60 80 100 0 100 200 300 400 500 % 15N EC

Correlation EC & 15_{N - Catchment 1}

2016

Correlation EC & 15N - Catchment 1 2016

Lineair (Correlation EC & 15N - Catchment 1 2016) y = 8,9758x + 41,675 r² = 0,34003 0 50 100 150 200 250 300 350 400 450 0 10 20 30 40 EC C:N ratio Correlation between C:N ratio and EC - Catchment 1 2016 Correlation between C:N ratio and EC - Catchment 1 2016

Lineair (Correlation between C:N ratio and EC - Catchment 1 2016) y = -2,5767x + 28,415 r² = 0,03862 -10,00 0,00 10,00 20,00 30,00 40,00 0 2 4 6 8 15N pH

Correlation pH & 15N - Catchment 2 2016

Correlation pH & 15N - Catchment 2 2016

Lineair (Correlation pH & 15N - Catchment 2 2016)

(13)

Figure 11. Correlation between EC and 15_{N in catchment 2, 2016. r}2_{= 0,19 and r = 0,43.}

Figure 12. Correlation between C:N ratio and EC in catchment 2, 2016. r2_{= 0,44 and r = 0,66.}

For both catchments three different slope types were distinguished: even, convex and concave. In catchment 1 the highest 15

N retention was found in the convex samples. There was a weak correlation between hill slope and 15

N, with a Pearson’s r of 0,03 and a determination coefficient of 0,17. In the second catchment, the highest 15

N retention was found in the concave hillslopes. There was a weak to moderate correlation between hillslopes, the Pearson’s r was 0,24 and the determination coefficient was 0,06. Overall the differences in soil moisture content, pH and EC were insignificant between catchment 1 and 2 (Appendix A.1). y = 8,5069x + 71,373 r² = 0,19559 0 50 100 150 200 250 300 350 400 450 0 5 10 15 20 25 EC C:N ratio

Correlation between C:N ratio and EC - Catchment 2 2016

Correlation between C:N ratio and EC - Catchment 2 2016

Lineair (Correlation between C:N ratio and EC - Catchment 2 2016) y = 0,0522x + 5,2879 r² = 0,1879 -5,00 0,00 5,00 10,00 15,00 20,00 25,00 30,00 35,00 40,00 0 100 200 300 400 500 15N EC

Correlation EC & 15_{N - Catchment 2}

2016

Correlation EC & 15N - Catchment 2 2016

Lineair (Correlation EC & 15N - Catchment 2 2016)

(14)

BSc Thesis – The long-term fate of deposited 15N in a coniferous forest in Alptal, Switzerland Quirine Cohen – 10516794 – 05 May 2018 – University of Amsterdam Figure 13. Mean %15_{N recovery per type of hillslope catchment 1 & catchment 2, 2016.}

Catchment 1: r2_{= 0,03 and r = 0,17. Catchment 2: r}2 _{= 0,24 and r = 0,06.}

Table 1. %15_{N recovery and C:N ratio per catchment per horizon (Schleppi et al., 2004).}

1997 data from (Schleppi et al., 1999); 2002/2003 data from (Providoli et al. 2005); SEM = standard error of the mean; 2016 for low N samples LFH horizon n = 17 A horizon n = 22, B horizon n = 22; for high N samples LFH horizon n = 12, A horizon n = 17, B horizon n = 14. 23,86 35,28 23,27 13,5 9,19 20,86 0,00 5,00 10,00 15,00 20,00 25,00 30,00 35,00 40,00

Even Convex Concave

% 15N r ec ove ry

Mean %15_{N recovery per hillslope}

Catchment 1 & 2 - 2016

Catchment 1 (2016) Catchment 2 (2016)

Low N plot - Catchment 1

Organic Soil Lfh % 15 N C:N ratio 2002/2003 2016 2016 SEM 2002/2003 2016 21,4 49 8,29 21 27,8 Mineral Soil A horizon B horizon 7,1 16,8 31,6 18,2 7,29 7,45 17,1 17,7 18,1 15,2 Total soil 45,3 98,7

High N plot - Catchment 2

Organic Soil Lfh % 15 N C:N ratio 2002/2003 2016 2016 SEM 1997 2002 2016 2016 SEM 13 20,2 3,0 21,7 2,0 Mineral Soil A horizon B horizon 42 21 18,3 8,8 2,5 3,0 17,7 15,9 1,9 4,3 Total soil 76 47,3 20,2 17,18 17,9 3,9

(15)

Table 2. Soil moisture content, pH, EC and slope gradient per horizon for each catchment in 2016.

Low N plot – Catchment 1

Organic Soil Lfh Soil moisture content SEM pH SEM EC SEM 72,7 3,49 2,97 3,12 4,3 0,36 0,33 0,28 291,5 24,7 Mineral Soil A horizon B horizon 69,8 50,4 4,1 4,6 148,9 148,4 24,7 19,6

High N plot – Catchment 2

Organic Soil Lfh Soil moisture content SEM pH SEM EC SEM 62 4,17 3,6 3,9 4,2 0,21 0,19 0,17 295,6 22 Mineral Soil A horizon B horizon 61,7 50 4,3 5,0 215,1 173 19,9 18,3

(16)

Discussion

Which soil layer (Lfh, A or B horizon) is the major 15

N sink, 13 and 19 years after 15 N deposition?

In the first catchment, the total 15

N recovery was 98% and the Lfh horizon was the major 15 N sink where 49% 15

N was recovered. The recovery was much higher than in 2002/2003 when in total 45.3% was recovered, especially in the Lfh and A horizon (table 2). Based on the results of Tietema et al., (unpublished results) we hypothesized that over this long time period, most of the 15

N would have leached from the organic horizon into the mineral soil. However, the organic soil layers remained the dominant 15

N sink after 19 years. When we combined the 15

N recovery levels of the A and B horizon (0 - 50cm), the 15

N recovery in the mineral soil would only be slightly higher (49.8%), but not significant.

The high 15

N recovery in the organic layer is possibly linked to an increase in N due to N immobilisation during decomposition or through new litter fall (Providoli et al. 2005). The majority of 15

N in 2002/2003 was recovered in the above ground vegetation and therefore could have decayed and released 15

N into the organic soil horizon (Providoli et al., 2005, 2006; Schleppi et al., 2004). This is in contrary with Persson & Wiren (1995) where a higher mineralisation rate was found in the Lfh horizon with a low C:N ratio. In order to conclude if the increase in 15

N retention in the organic soil layer is a result of 15

N litter fall, 15

N recovery should be measured in the above ground vegetation and in the roots in the catchment. Since 15_{N was almost fully recovered, there was no N leaching in the catchment. 24.5% more}15_N has leached in the A horizon in 2016 apropos of 2002/2003 and it is expected that 15

N will leach further into the B horizon in the future.

In the second catchment the total 15

N recovery has decreased over time from 76% to 47.3%. The major 15_{N sink in catchment 2 was the Lfh horizon (20.2%). However, when}15_{N recovery} in the A and B horizon was combined, the mineral soil retained more 15

N (27.1%) in comparison with the Lfh horizon. Similar to catchment 1, there was an increase since 1997 in 15

N retention in the organic soil horizon of 7.2%. This is also linked to 15

N immobilisation during decomposition and incorporation of new 15

N litter in the organic soil Since 28.7% of the 15

N recovery was not accounted for, it is expected that this has leached out of the mineral soil layer or that a larger part remained available in the vegetation pool. In Schleppi et al. (2004) it is proposed that N leaching is caused by three different mechanisms. NO3

-

leaching from precipitation, flushing of NO3

-

after storage in the soil pores or flushing produced by nitrification. This was also found in Schleppi et al. (1998b, 1999) when leaching directly came from the treatment of N addition and was at least partially hydrologically driven. This meant that at that moment there was no sufficient proof that the ecosystem was N-saturated (Schleppi et al., 1999). Furthermore, N leaching was found in 2005 by Providoli et al., who measured the highest peak in NO3

- _{leaching in winter after snowmelt. Moreover, in previous} research in acid forest soils during wintertime, the B horizon was the major source of NO3

-, leading to further NO3

-

leaching (Persson and Wirén 1995). Likewise in 2005 immediate leaching after 15

N addition was measured and is associated with either runoff or water saturation (Creed & Band, 1998; Hagedorn et al., 2001; Providoli et al., 2005; Schleppi et al., 2004). Also, the increase in NO3

-

leaching cannot be a result of the postponed discharge of N accumulated in preceding years, due to the disappearing 15

N in the research in Schleppi et al. (2004).

(17)

What is the effect of continuous N addition on 15

N retention in the soil?

In catchment 2, there has been a continuous increased N addition since 1997, leading to a total addition of 42 kg N ha-1

yr-1, in comparison with a continuous ambient N addition of 12 kg N ha-1

yr-1

. The main effect of the increased N addition is that more N seemed to have leached out of the soil when compared with ambient N addition. The total 15

N recovery in catchment 2 (45.3%) is much lower in comparison with catchment 1 (98.8%). When 15

N retention was compared per horizon, all horizons showed higher 15

N recovery levels in catchment 1 compared to catchment 2. The difference between the two catchments can be explained by increased 15

N leaching, due to N saturation or more N is stored in other N pools. Unfortunately we did not investigate 15

N retention in those other possible N pools. When comparing C:N ratio’s between catchments, it was expected that despite increased N leaching in catchment 2, the increased N deposition would result in lower soil C:N ratio’s, assuming C content remained the same. A significant difference was measured in the organic layer (p < 0.000002), with indeed a lower C:N ratio in the second catchment. Overall, the C:N ratio in catchment 2 (17.9) was lower than in catchment 1 (19.7). When comparing %C content between horizons in both catchments, it was found that the values indeed remained fairly the same. Between catchment 1 and 2, the Lfh horizon had a %C of 39.93% and 42.74% respectively. The average %C in the A horizon of catchment 1 was 19.7% and in the B horizon 6.21%. In the second catchment, the average %C in the A horizon was 21.49% and in the B horizon 6.7%. Therefore it is concluded that C content remains the same and a continuous N addition leads to a lower C:N ratio. In Schleppi et al. (2004) it is hypothesized that the declining C:N ratio is increasingly limiting the soils ability to immobilise N from atmospheric deposition. This could be an explanation of the lower 15

N retention and higher N saturation rate in catchment 2 in comparison with catchment 1.

How is 15

N retention influenced by soil moisture content, pH, EC and slope gradient?

In catchment 1 15

N decreases per horizon, (fig. 2) and the soil moisture content decreased per horizon (table 1). It was expected that this was a consequence of the low permeability of the gleyic hapzol, which does not allow the water to penetrate the soil (Christopherson 2013). The pH decreases from Lfh to A horizon and then increases from A to B horizon. As mentioned before, with an increasing pH the microbes in the soil can become more active, decreasing the C:N ratio. A higher pH could therefore lead to more 15

N retention in the soil. Since the pH was lower in the Lfh horizon, another possibility for more 15

N retention is through N leaching. In the research by Koopmans, Tietema, and Boxman (1996) in sandy soils, a low pH showed lower N storage capacities which led to further N leaching by water. Although very weak, a positive correlation was found between pH and 15

N. This corresponds with the decreasing C:N ratio per horizon.

Soil EC can be influenced by different factors such as soil moisture content, ion concentration, the amount and type of clay and the soil bulk density (Zhang & Wienhold 2002). The EC decreased from the organic soil horizon and fairly remained constant from the A to B horizon. When EC increases, the concentration of inorganic N increases due to mineralization (Sánchez -Monedero et al., 2001; Zhang & Wienhold, 2002). A moderate positive correlation was found between EC and 15

N; when EC increased, more 15

N was recovered. This increase in EC is correlated to the increased concentrations of nutrients. The slope gradients observed in the catchment were even, convex and concave. The highest 15

N retention was measured in samples that were taken from a convex area. This could be a consequence of accumulation of leached N from higher parts in the catchment. However, these differences were insignificant and no correlation was found between hill slope and 15

(18)

In the second catchment 15

N also decreases per horizon (fig. 3). The soil moisture decreases per horizon but is only significant from Lfh to B and the A to B horizon. In this catchment, it is also expected that this is a consequence of the low permeability of the soil. Moreover, pH increases significantly from the Lfh to B horizon and the A and B horizon. This can also be seen in the correlation coefficients, where a moderate to strong correlation was found. This increase in pH might lead to less 15

N retention due to more microbial activity. The EC measured per horizon, decreases with depth, similar to catchment 1, and is significant from the organic soil horizon to the A and B horizon. However, as in catchment 1, a moderate positive correlation was found; where a higher EC was measured, more 15

N was retained. This positive correlation is a direct effect of a higher nutrient concentration (Sánchez -Monedero et al. 2001). The positive correlation could also be a consequence of the lower C:N ratio measured at catchment 2. When C:N ratio decreases, N mineralization increases, which can lead to higher EC (Zhang & Wienhold 2002). A moderate correlation was found between C:N ratio and EC in catchment 2, with Pearson’s r of 0,44, which could have influenced the increasing EC. However, the correlation coefficients between EC and 15

N and EC and C:N ratio were almost equal. The highest 15

N retention was measured in the concave areas, but was not significant and no correlation was found between 15

N retention and hill slope. Since in the first catchment most 15

N was found in convex areas, further research needs to be done on the effect of slope gradient on 15_{N retention.}

Conclusion

In both catchments the major 15

N sink was the organic soil horizon. However, when A and B horizons were combined, 15

N retention was higher in the mineral soil (0-50 cm) in comparison with the organic soil. This was in line with previous research done by Tietema et al. (unpublished results). In the first catchment almost 100% of the added 15_{N was recovered in} the soil after more than 20 years, in contrary to previous years. This increase in 15

N retention in the soil can be linked to immobilisation during decomposition, abiotic absorption and the incorporation of 15

N litter in the soil. In the second catchment, where high levels of N were added (42 kg N ha-1

yr-1

) less 15

N was recovered in the soil in comparison to catchment 1. This might confirm that with higher N addition more of the added 15_{N was leached out of the} system due to N saturation or remained available in the vegetation pool. In catchment 1, a moderate correlation was found between electrical conductivity and 15

N. This corresponds with the high 15

N retention. In the second catchment, a negative correlation was found between pH and 15

N. This could explain the low 15

N retention in the soil, since high pH increases microbial activity. Finally, in both catchments the influence of the slope on the 15

N retention was measured, but showed an insignificant correlation.

(19)

References

Aber, John D, Knute J Nadelhoffer, Paul Steudler, and Jerry M Melillo. 1989. “Nitrogen Saturation in Northern Forest Ecosystems Published by : Oxford University Press on Behalf of the American Institute of Biological Sciences Stable.” 39(6): 378–86.

Burt, J.E., G.M. Barber, and D.L. Rigby. 2009. Elementary Statistics for Geographers. New York: The Guilford Press.

Christopherson, R.W. 2013. Elemental Geosystems. 7th ed. Pearson.

Creed, I F, and L E Band. 1998. “Exploring Functional Similarity in the Export of Nitrate-N from Forested Catchments : A Mechanistic Modeling Approach.” Water Resources

Research 34(11): 3079–93.

Dise, N.B., and R.F. Wright. 1995. “Nitrogen Leaching from European Forests in Relation to Nitrogen Deposition.” Forest Ecology and Management 71(94): 153–61.

Dise, N.B., and R F Wright. 1992. The NITREX Project.

Goodale, C.L. 2016. “Multiyear Fate of a 15N Tracer in a Mixed Deciduous Forest:

Retention, Redistribution, and Differences by Mycorrhizal Association.” Global Change

Biology 23(2): 867–80.

http://onlinelibrary.wiley.com/resolve/doi?DOI=10.1111%2Fgcb.13483.

Hagedorn, F., Maurer, S., Egli, P., Blaser, P.E., Bucher, J.B. and Siegwolf, R. 2001. “Carbon Sequestration in Forest Soils : Effects of Soil Type , Atmospheric CO 2 Enrichment , and N Deposition.” Science (December).

Koopmans, Author C J, D Van Dam, A Tietema, and J M Verstraten. 1997. “International Association for Ecology Natural 15N Abundance in Two Nitrogen Saturated Forest Ecosystems.” Ecology 111(4): 470–80.

Koopmans, C . J ., A. Tietema, and A.W. Boxman. 1996. “The Fate of 15N Enriched

Throughfall in Two Coniferous Forest Stands at Different Nitrogen Deposition Levels.” 34(1): 19–44.

Koopmans, C J, W C Lubrecht, and A Tietema. 1995. “Nitrogen Transformations in Two Nitrogen Saturated Forest Ecosystems Subjected to an Experimental Decrease in Nitrogen Deposition.” Plant and Soil 175(2): 205–18.

(20)

Persson, T., and A. Wirén. 1995. “Nitrogen Mineralization and Potential Nitrification at Different Depths in Acid Forest Soils.” Plant and Soil 168(1): 55–65.

Providoli, I. Bugmann, H, Siegwolf, R., Buchmann, N., Schleppi, P. 2005. “Flow of

Deposited Inorganic N in Two Gleysol-Dominated Mountain Catchments Traced with 15NO3 - And 15NH 4+.” Biogeochemistry 76(3): 453–75.

Providoli, I. Bugmann, H, Siegwolf, R., Buchmann, N. and Schleppi, P. 2006. “Pathways and Dynamics of 15 NO3- and 15 NH4+ Applied in a Mountain Picea Abies Forest and in a Nearby Meadow in Central Switzerland.” Soil Biology & Biochemistry 38: 1645–57. Sánchez -Monedero, M.A., A. Roig, C. Paredes, and M.P. Bernal. 2001. “Nitrogen

Transformation during Organic Waste Composting by the Rutgers System and Its Effects on pH, EC and Maturity of the Composting Mixtures.” Bioresource Technology 78: 301–8.

Schleppi, P. Bucher-Wallin, I., Siegwolf, R., Saurer, M., Muller, N. and Bucher J.B. 1999. “Simulation of Increased Nitrogen Deposition To a Montane.” Methods: 129–34. Schleppi, P., Muller, N., Feyen, H., Papritz, A., Bucher, J.B. and Flühler, H. 1998. “Nitrogen

Budgets of Two Small Experimental Forested Catchments at Alptal, Switzerland.”

Forest Ecology and Management 101(1–3): 177–85.

Schleppi, Patrick, Frank Hagedorn, and Isabelle Providoli. 2004. “Nitrate Leaching from a Mountain Forest Ecosystem with Gleysols Subjected to Experimentally Increased N Deposition.” Water, Air, and Soil Pollution: Focus 4(2–3): 453–67.

Templer, P H, Mack, M.C., Chapin, F.S., Christenson, L.M., Crompton, J.E., Crook, H.D. … D'Antonio, C.M. 2012. “Sinks for Nitrogen Inputs in Terrestrial Ecosystems: A Meta-Analysis of 15 N Tracer Field Studies.” Ecology 93(8): 1816–29.

Wessel, Wim W., Albert Tietema, and Andries W. Boxman. 2013. “The Fate of 15NH4+ Labeled Deposition in a Scots Pine Forest in the Netherlands under High and Lowered NH4+ Deposition, 8 Years after Application.” Biogeochemistry 113(1–3): 467–79. Zhang, R., and B.J. Wienhold. 2002. “The Effect of Soil Moisture on Mineral Nitrogen , Soil

(21)

Appendix

A.1. p –values

Variable Catchment 1 p value Catchment 2 p value Comparison p value

C1 C2 Nitrogen LFH - A 0.1218 LFH - A 0.6403 Lfh C1 – Lfh C2 0.0239 Nitrogen LFH - B 0.0078 LFH - B 0.0142 A C1 - A C2 0.3467 Nitrogen A -B 0.2025 A - B 0.0262 B C1 - B C2 0.0678 C1 C2 C:N LFH - A 0.0000 LFH - A 0.0000 Lfh C1 – Lfh C2 2.0E-6 C:N LFH - B 0.0000 LFH - B 0.0000 A C1 - A C2 0.9993 C:N A -B 0.0566 A - B 0.0306 B C1 - B C2 0.9981 C1 C2 SM LFH - A 0.5260 LFH - A 0.9570 Lfh C1 – Lfh C2 0.3800 SM LFH - B 0.0000 LFH - B 0.0430 A C1 - A C2 0.5290 SM A -B 0.0000 A -B 0.0340 B C1 - B C2 1.0000 C1 C2 pH LFH - A 0.6785 LFH - A 0.6785 Lfh C1 – Lfh C2 0.9999 pH LFH - B 0.5069 LFH - B 0.5069 A C1 - A C2 1.0000 pH A -B 0.2475 A - B 0.2475 B C1 - B C2 0.9060 C1 C2 EC LFH - A 0.0002 LFH - A 0.0109 Lfh C1 – Lfh C2 1.0000 EC LFH - B 0.0001 LFH - B 0.0001 A C1 - A C2 0.3310 EC A -B 0.9882 A - B 0.0837 B C1 - B C2 0.9490 C1 C2 Slope LFH - A 0.9115 LFH - A 0.7798 Slope LFH - B 0.2141 LFH - B 1.0000 Slope A -B 0.2504 A - B 0.8087

(22)

A.2. Correlation graphs

Catchment 1

Figure 1. Correlation between pH and 15_{N catchment 1, 2016. r}2_{= 0,026 and r = 0,40.}

Figure 2. Correlation between moisture content and 15_{N catchment 1, 2016. r}2_{= 0.02 and r = 0,36.}

y = 2,4616x + 12,183 r² = 0,026 0 10 20 30 40 50 60 70 0 1 2 3 4 5 6 7 8

Correlation pH & 15_{N - Catchment 1}

2016 Correlation pH & 15N - Catchment 1 Lineair (Correlation pH & 15N - Catchment 1) y = 0,0879x + 11,805 r² = 0,01635 0 5 10 15 20 25 30 35 40 45 0 20 40 60 80 100

Correlation moisture content & 15_{N - Catchment 1}

2016

Correlation Moisture Content & 15N - Catchment 1 2016

Lineair (Correlation Moisture Content & 15N - Catchment 1 2016)

(23)

Catchment 2

Figure 3. Correlation between moisture content and 15_{N catchment 2, 2016. r}2_{= 0,00025 and r = 0,016.}

A.3. MatLab scripts

Catchment 1 – Compare 15N retention per horizon

% %% 15N recovery Catchment 1 Catchment = 'Catchment1_Nrecovery.xlsx' Catchment1_Nrecovery = xlsread(Catchment) N_Lfh_row = Catchment1_Nrecovery(1:17,29); N_A_row = Catchment1_Nrecovery(18:39,29); N_B_row = Catchment1_Nrecovery(40:61,29); N_LFH = N_Lfh_row' N_A = N_A_row' N_B = N_B_row' N = [N_LFH N_A N_B] group = Catchment1_Nrecovery(1:61,33)

[h, p, stats] = anovan(N, group)

[c3,m3] = multcompare(stats,'alpha', 0.05,'ctype','lsd') y = -0,0094x + 15,444 r² = 0,00025 -5 0 5 10 15 20 25 30 35 40 0 20 40 60 80 100

Correlation moisture content and 15_{N - Catchment 2 2016}

Correlation moisture content and 15N - Catchment 2 2016

Lineair (Correlation moisture content and 15N - Catchment 2 2016)

(24)

Catchment 1 – Comparing C:N ratio per horizon

% %% C:N ratio Catchment 1 Catchment1 = 'Catchment1_CNRATIO.xlsx' Catchment1_CNRatio = xlsread(Catchment1) CN_Lfh_C1 = Catchment1_CNRatio(3:18,24); CN_A_C1 = Catchment1_CNRatio(19:39,24); CN_B_C1 = Catchment1_CNRatio(40:59,24); CN_LFH_C1 = CN_Lfh_C1' CN_A_C1 = CN_A_C1' CN_B_C2 = CN_B_C1' N1 = [CN_LFH_C1 CN_A_C1 CN_B_C2] group1 = Catchment1_CNRatio(3:59,25)

[h, p, stats] = anovan(N1, group1)

[c_CN_C1,m_CN_C1] = multcompare(stats,'alpha', 0.05,'ctype','lsd')

Catchment 1 – Comparing soil moisture per horizon

% %% Moisture Content - Catchment 1 Catchment1 = 'Catchment1_Nrecovery.xlsx' Catchment1_MC = xlsread(Catchment1) MC_Lfh_C1 = Catchment1_MC(2:17,31); MC_A_C1 = Catchment1_MC(18:39,31); MC_B_C1 = Catchment1_MC(40:59,31); MC_Lfh_C1 = MC_Lfh_C1' MC_A_C1 = MC_A_C1' MC_B_C1 = MC_B_C1' N_MC_1 = [MC_Lfh_C1 MC_A_C1 MC_B_C1] group_MC_1 = Catchment1_MC(2:59,32)

[h_MC_1, p_MC_1, stats_MC_1] = anovan(N_MC_1, group_MC_1)

(25)

Catchment 1 – Comparing pH per horizon

Catchment = 'Catchment1_CNRATIO.xlsx' Catchment1_pH = xlsread(Catchment) % pH_LFH_C1 = Catchment1_pH(2:13,34); pH_A_C1 = Catchment1_pH(14:27,34); pH_B_C1 = Catchment1_pH(28:47,34); pH_LFH_C1 = pH_LFH_C1' pH_A_C1 = pH_A_C1' pH_B_C1 = pH_B_C1' N_ph_1 = [pH_LFH_C1 pH_A_C1 pH_B_C1] group_pH_C1 = Catchment1_pH(2:47,35)

[h_pH_1, p_pH_1, stats_pH_1] = anovan(N_ph_1, group_pH_C1)

[c_pH_1,m_pH_1] = multcompare(stats_pH_1,'alpha', 0.05,'ctype','lsd')

Catchment 1 – Comparing EC per horizon

Catchment = 'Catchment1_CNRATIO.xlsx' Catchment1_EC = xlsread(Catchment) EC_LFH_C1 = Catchment1_EC(2:13,36); EC_A_C1 = Catchment1_EC(14:25,36); EC_B_C1 = Catchment1_EC(26:44,36); EC_LFH_C1 = EC_LFH_C1' EC_A_C1 = EC_A_C1' EC_B_C1 = EC_B_C1'

N_EC_1 = [EC_LFH_C1 EC_A_C1 EC_B_C1] group_EC_1 = Catchment1_EC(2:44,38)

[h_EC_1, p_EC_1, stats_EC_1] = anovan(N_EC_1, group_EC_1)

[c_EC_1,m_EC_1] = multcompare(stats_EC_1,'alpha', 0.05,'ctype','lsd')

Catchment 1 – Comparing slope per horizon

Catchment = 'Catchment1_Nrecovery.xlsx' Catchment1_Nrecovery = xlsread(Catchment); Recovery1 = Catchment1_Nrecovery(2:62,29) slope1 = Catchment1_Nrecovery(2:62,34)

[pslope1,tblslope1,statsslope1] = anovan(Recovery1, slope1)

(26)

Catchment 2 – Comparing 15N retention per horizon

% %% 15N recovery Catchment 2 Catchment2 = 'Catchment2_Nrecovery.xlsx' Catchment2_Nrecovery = xlsread(Catchment2) N_Lfh_row2 = Catchment2_Nrecovery(1:12,29); N_A_row2 = Catchment2_Nrecovery(13:29,29); N_B_row2 = Catchment2_Nrecovery(30:43,29); N_LFH2 = N_Lfh_row2' N_A2 = N_A_row2' N_B2 = N_B_row2' N2 = [N_LFH2 N_A2 N_B2] group2 = Catchment2_Nrecovery(1:43,30)

[h, p, stats] = anovan(N2, group2)

[c3,m3] = multcompare(stats,'alpha', 0.05,'ctype','lsd')

Catchment 2 – Comparing C:N ratio per horizon

% %% C:N ratio Catchment 2 Catchment2 = 'Catchment2_CNRATIO.xlsx' Catchment2_CNRatio = xlsread(Catchment2) CN_Lfh_C2 = Catchment2_CNRatio(1:12,23); CN_A_C2 = Catchment2_CNRatio(13:25,23); CN_B_C2 = Catchment2_CNRatio(26:40,23); CN_LFH_C1 = CN_Lfh_C2' CN_A_C2 = CN_A_C2' CN_B_C2 = CN_B_C2' N_C2 = [CN_LFH_C1 CN_A_C2 CN_B_C2] group2 = Catchment2_CNRatio(1:40,24)

[h, p, stats] = anovan(N_C2, group2)

(27)

Catchment 2 – Comparing soil moisture per horizon

% %% Moisture Content - Catchment 2 Catchment2 = 'Catchment2_Nrecovery.xlsx' Catchment2_MC = xlsread(Catchment2) MC_Lfh_C2 = Catchment2_MC(2:13,36); MC_A_C2 = Catchment2_MC(14:29,36); MC_B_C2 = Catchment2_MC(30:44,36); MC_Lfh_C2 = MC_Lfh_C2' MC_A_C2 = MC_A_C2' MC_B_C2 = MC_B_C2' N_MC_2 = [MC_Lfh_C2 MC_A_C2 MC_B_C2] group_MC_2 = Catchment2_MC(2:44,37)

[h_MC_2, p_MC_2, stats_MC_2] = anovan(N_MC_2, group_MC_2)

[c_MC_2,m_MC_2] = multcompare(stats_MC_2,'alpha', 0.05,'ctype','lsd')

Catchment 2 – Comparing pH per horizon

Catchment = 'Catchment2_Nrecovery.xlsx' Catchment2_pH = xlsread(Catchment) pH_LFH_C2 = Catchment2_pH(2:10,32); pH_A_C2 = Catchment2_pH(11:22,32); pH_B_C2 = Catchment2_pH(23:37,32); pH_LFH_C2 = pH_LFH_C2' pH_A_C2 = pH_A_C2' pH_B_C2 = pH_B_C2' N_pH_2 = [pH_LFH_C2 pH_A_C2 pH_B_C2] group_ph_2 = Catchment2_pH(2:37,33)

[h_pH_2, p_pH_2, stats_pH_2] = anovan(N_pH_2, group_ph_2)

(28)

Catchment 2 – Comparing EC per horizon

Catchment = 'Catchment2_Nrecovery.xlsx' Catchment2_EC = xlsread(Catchment) EC_LFH_C2 = Catchment2_EC(2:10,34); EC_A_C2 = Catchment2_EC(11:21,34); EC_B_C2 = Catchment2_EC(22:35,34); EC_LFH_C2 = EC_LFH_C2' EC_A_C2 = EC_A_C2' EC_B_C2 = EC_B_C2'

N_EC_2 = [EC_LFH_C2 EC_A_C2 EC_B_C2] group_EC_2 = Catchment2_EC(2:35,35)

[h_EC_2, p_EC_2, stats_EC_2] = anovan(N_EC_2, group_EC_2)

[c_EC_2,m_EC_2] = multcompare(stats_EC_2,'alpha', 0.05,'ctype','lsd')

Catchment 2 – Comparing slope per horizon

Catchment = 'Catchment1_Nrecovery.xlsx' Catchment2_Nrecovery = xlsread(Catchment); Recovery2 = Catchment2_Nrecovery(63:105,29) slope2 = Catchment2_Nrecovery(63:105,34)

[pslope2,tblslope2,statsslope2] = anovan(Recovery2, slope2)

[Slope_C2 MeanNslope2] = multcompare(statsslope2,'alpha', 0.05) %

Catchment 1 & 2 – 15N retention

%15N recovery per horizon

%ANOVA to discover difference between catchments. %% Catchment 1 initialisation Catchment = 'Catchment1_Nrecovery.xlsx' Catchment1_Nrecovery = xlsread(Catchment) N_Lfh_row = Catchment1_Nrecovery(2:18,30); N_A_row = Catchment1_Nrecovery(19:40,30); N_B_row = Catchment1_Nrecovery(41:62,30); N_LFH = N_Lfh_row'; N_A = N_A_row'; N_B = N_B_row' ;

(29)

%% Catchment 2 initialisation Catchment2 = 'Catchment2_Nrecovery.xlsx' Catchment2_Nrecovery = xlsread(Catchment2); N_Lfh_row2 = Catchment2_Nrecovery(1:12,29); N_A_row2 = Catchment2_Nrecovery(13:29,29); N_B_row2 = Catchment2_Nrecovery(30:43,29); N_LFH2 = N_Lfh_row2'; N_A2 = N_A_row2'; N_B2 = N_B_row2';

%% ANOVA Catchment 1 & 2 % LFH horizon

N = [N_LFH N_LFH2]

group = Catchment1_Nrecovery(1:29,42) [h, p, stats] = anovan(N, group)

[c3,m3] = multcompare(stats,'alpha', 0.05,'ctype','lsd') %% Conclusion

% There is a significant difference between the 2 catchments in horizon % LFH.

%% ANOVA Catchment 1&2 % A horizon

N = [N_A N_A2]

% There is no significant difference between the 2 catchments in horizon A.

%% ANOVA Catchment 1&2 % B horizon

N = [N_B N_B2]

(30)

Catchment 1 & 2 – C:N retention

% %% EC - Catchment 1 & 2 % 1 = Catchment 1; Lfh horizon % 2 = Catchment 1; A horizon % 3 = Catchment 1; B horizon % 4 = Catchment 2; Lfh horizon % 5 = Catchment 2; A horizon % 6 = Catchment 2; B horizon Catchment = 'Catchment2_CNRATIO.xlsx' Catchment1_2_CN = xlsread(Catchment); C1_C2_CN = Catchment1_2_CN(2:98,25) groups_CN= Catchment1_2_CN(2:98,26)

[pC1_C2_CN,tblC1_C2_CN,statsC1_C2_CN] = anovan(C1_C2_CN, groups_CN)

[pvalC1_C2_CN meanC1_C2_CN] = multcompare(statsC1_C2_CN,'alpha', 0.05) %

Catchment 1 & 2 – Soil moisture content

% % %% MC - Catchment 1 & 2 % % 1 = Catchment 1; Lfh horizon % % 2 = Catchment 1; A horizon % % 3 = Catchment 1; B horizon % % 4 = Catchment 2; Lfh horizon % % 5 = Catchment 2; A horizon % % 6 = Catchment 2; B horizon % Catchment = 'Catchment1_Nrecovery.xlsx' Catchment1_2_MC = xlsread(Catchment); C1_C2_MC = Catchment1_2_MC(2:102,31) groups_MC= Catchment1_2_MC(2:102,32)

[pC1_C2_MC,tblC1_C2_MC,statsC1_C2_MC] = anovan(C1_C2_MC, groups_MC)

(31)

Catchment 1 & 2 – pH % %% pH - Catchment 1 & 2 % 1 = Catchment 1; Lfh horizon % 2 = Catchment 1; A horizon % 3 = Catchment 1; B horizon % 4 = Catchment 2; Lfh horizon % 5 = Catchment 2; A horizon % 6 = Catchment 2; B horizon Catchment = 'Catchment1_CNRATIO.xlsx' Catchment1_pH = xlsread(Catchment); pH_C1_C2 = Catchment1_pH(2:83,34) groups_CN = Catchment1_pH(2:83,35)

[pC1_C2_ph,tblC1_C2_ph,statsC1_C2_ph] = anovan(pH_C1_C2, groups_CN)

[pvalC1_C2_ph meanC1_C2_ph] = multcompare(statsC1_C2_ph,'alpha', 0.05) %

Catchment 1 & 2 – EC

% %% 15N recovery - Catchment 1 & 2 % 1 = Catchment 1; Lfh horizon % 2 = Catchment 1; A horizon % 3 = Catchment 1; B horizon % 4 = Catchment 2; Lfh horizon % 5 = Catchment 2; A horizon % 6 = Catchment 2; B horizon Catchment = 'Catchment1_CNRATIO.xlsx' Catchment2_EC = xlsread(Catchment); EC_C1_C2 = Catchment2_EC(2:78,36) groups_EC = Catchment2_EC(2:78,37)

[pC1_C2_EC,tblC1_C2_EC,statsC1_C2_EC] = anovan(EC_C1_C2, groups_EC)

[pvalC1_C2_EC meanC1_C2_EC] = multcompare(statsC1_C2_EC,'alpha', 0.05) %