Assessing the long-term influence of active blowout processes on soil quality in the Grey dunes of Eldorado, Terschelling

(1)

Assessing the long-term

influence of active blowout

processes on soil quality in the

Grey dunes of Eldorado,

Terschelling

BACHELORPROJECT EARTH SCIENCES

Pauline Martens

First Supervisor: Dr. Annemieke Kooijman Second Supervisor: Dr. John van Boxel Bsc Earth Sciences Thesis Student number: 10485228

(2)

2

Abbreviations

A Stable blowout

B Recently stabilized blowout C Carbon

EC Electrical conductivity

GIS Geographic information system N Nitrogen

NAP Amsterdam Ordenance Datum ‘Normaal Amsterdams peil’ OM Organic matter

P Phosporus

PAS Programma aanpak stikstof SOM Soil organic matter

(3)

3

Abstract

High atmospheric nitrogen deposition has been causing eutrophication and acidification of soils in Grey dunes, negatively affecting the habitat and its biodiversity. Reactivation of blowouts can be used to counteract this. To assess this measure it is important to know the effects of naturally active blowouts in the long run. This research aims to assess the long-term impact of active blowout processes on soil development, eutrophication and acidification of the soil in a natural reference situation. The outcome can be used to determine the suitability of reactivation as a management measure. The study took place on Terschelling in the lime-poor Grey dunes of Eldorado, where a recently stabilized blowout was compared to a fully stabilized blowout. Results were produced based on fieldwork, laboratory analysis of soil samples and GIS analysis to assess the effects of blowout activity on soil quality. The findings are that naturally active blowout processes have resulted in a lower degree of soil development in the recently stabilized blowout and have counteracted eutrophication. This was indicated by significantly lower N contents in the recently stabilized blowout and the significant difference in soil development, bulk density and bare sand coverage. This indicates a positive effect of active blowout processes on the soil. The lack of a significant difference in pH excluded that naturally active blowout processes had counteracted acidification on the long-term. Based on the findings, the overall effect appears to be positive, indicating that reactivation would be a good management measure.

Samenvatting

Verhoogde stikstof depositie resulteert in verzuring en vermesting van bodems in de Grijze duinen. Dit leidt tot vergrassing en een algehele daling in biodiversiteit. Om dit proces tegen te gaan worden sinds de jaren negentig stuifkuilen gereactiveerd, met het idee meer dynamiek in het landschap te brengen. Om deze herstelmaatregel te beoordelen moet er ook kennis voor handen zijn over natuurlijk actieve verstuivingen en de effecten van verstuivingen op de bodem op de lange termijn. Het doel van deze studie is te beoordelen welke gevolgen actieve verstuivingen hebben op de bodemontwikkeling, verzuring en vermesting van de bodem. De resultaten kunnen gebruikt worden om vast te stellen of reactivatie een goede maatregel is. Het onderzoek heeft plaats gevonden op Terschelling, in de kalk-arme Grijze duinen van het gebied Eldorado. Hier is een recent gestabiliseerde verstuiving vergeleken met een eerder gestabiliseerde stuifkuil. De uitkomsten zijn gebaseerd op veldwerk, laboratorium en GIS analyses. De bevindingen zijn dat natuurlijke verstuivingen resulteren in minder bodemontwikkeling en lagere stikstof vermesting. De hoeveelheid stikstof in de recent gestabiliseerde kuil was significant minder, daarbij waren de bodem horizons dunner, was er meer kaal zand en een hogere bulk dichtheid van de bodem. Aan de andere kant was er geen significant verschil in pH, wat aangeeft dat verzuring van de bodem op de lange termijn niet tegen wordt gegaan door verstuivingen. Over het algemeen lijkt het effect van de actieve stuifkuil positief te zijn geweest, op zowel bodem als omgeving, dit geeft aan dat reactivatie een goede herstelmaatregel lijkt te zijn.

(4)

4

1. Introduction

Coastal dunes provide many important ecosystem services to humans, making them integral in the Wadden Islands where there is a relatively large proportion of coastal dune land area (Łabuz, 2015). Important services for these islands utilized by humans range from protection from storms and floods and opportunities for tourism and recreation (Łabuz, 2015). The coastal dunes also provide habitat to many different species of flora and fauna. Around 70 percent of all the national vegetation species and a number of endemic species are found in sand dunes (Veer & Kooijman, 1997). As a result of these functions it is important to protect and conserve the coastal dunes to maintain the ecosystem for future use.

However, over the course of the last century the coastal dunes have been significantly altered. High levels of atmospheric nitrogen deposition have resulted in both eutrophication and acidification in the dunes at various places along the Dutch coast. Eutrophication of the sand dunes results in encroachment of the dune surface with shrubs, grasses and mosses (Van der Meulen et al., 1996, Kooijman et al., 2005). This in turn limits the biodiversity of the system and reduces dynamic aeolian activity in the coastal dunes. The nutrient overload is mainly caused by the agricultural sector that uses a large amount of fertilizers for food production, but industry and fuel emissions have also contributed. Through monitoring and technological advances, nitrogen deposition is now decreasing again, after increasing exponentially during the 19th_{century (Hicks et al.,} 2011).

As a result of the high pressure put on the coastal dunes the habitat conditions have changed significantly. In an attempt to restore the dunes to their natural habitat conditions various measures were assessed over the period from 1991 to 1998 (Van der Meulen et al., 1996, Kooijman et al., 2005). One of the potential restoration measures was the reactivation of blowouts to increase aeolian activity and to restore the lime buffer of the soil, which would counteract acidification (Van der Meulen et al., 1996). This measure is part of the Dutch PAS agreement – Programma aanpak stikstof – to restore habitats within Natura 2000 areas and limit the effects of N deposition (Doekes et al., 2015). A goal of blowout reactivation during earlier research was that the newly surfaced sand would be transported by the wind and would serve as a substrate to facilitate pioneer vegetation (Van der Meulen et al., 1996), which would increase the biodiversity of the coastal dunes.

To assess whether blowout reactivation is a valuable restoration strategy, it is important to first study naturally active blowouts. By doing so it can be assessed if active blowout processes really do improve soil quality and consequently habitat conditions. During the active period of the blowout fresh sand was deposited in the surrounding area. This increased the pH, lowered the amounts of soil organic matter in the topsoil and created suitable habitat conditions for characteristic plant species. What was still unknown is how long the effects would last for, how significant they have been and whether the effects from naturally active blowouts differ from reactivated blowouts.

The aim of this research is to assess the impact of active blowout processes on the soil quality. The outcome can be used to argue if reactivation of stable blowouts in the Grey dunes is a suitable management measure. The main question that this study answers is: What is the long-term impact of active blowout processes on soil development

(6)

6 and acidification and eutrophication of the soil in a recently stabilized blowout - used as natural reference for reactivation - compared to an earlier stabilized blowout in the Grey dunes of Eldorado? The main hypothesis is that soil development is higher in the fully stabilized blowout and that acidification and eutrophication are counteracted in the recently stabilized blowout as a result of active blowout processes.

The first subquestion is: In what way has soil development been influenced by blowout activity? The answer is based on a characterization of the soil development in both blowouts given in chapter one and various soil properties. The second question that will be treated is: Does blowout activity counteract soil acidification on a long-term? This is mainly based on the pH dynamics of both blowouts as described in chapter two. The last sub question is: Is soil eutrophication counteracted by blowout activity on a long-term? The answer to this question is based on the carbon and nitrogen dynamics in chapter three.

After giving an overview of the theories and concepts relevant to this research and methodology, this report continues with the results. These are divided into three sections and follow the order of the subquestions as given above. The discussion further expounds on the findings and argues the significance and validity of the results. It then describes the differences between reactivated and naturally active blowouts. Finally, the conclusion readdresses the primary research question and summarizes the main findings.

1.1 Research area

The research took place in the area of Eldorado located on the Wadden island Terschelling (figure 1). This area was also studied during earlier research (Van der Meulen et al., 1996, Kooijman et al., 2005). Eldorado is located around 500m inland

from the beach zone and can be characterized as a fossilized beach zone with an irregular relief including a large amount of fossilized blowouts (Van der Meulen et al., 1996). The total area is around 80 ha and is located up to 20m above NAP. Until 1945 the area was very dynamic. While the dunes stabilized in the period afterwards, the general geomorphology remained largely the same (Van der Meulen et al., 1996 & Kooijman et al., 2005). This was mostly the result of the stabilization-focused policies at the time that prevented dynamic aeolian activity. These were implemented until 1980, to maintain the flood defensive function of the coastal dunes (Ibid.).

Figure 1. Overview map of the Wadden Island Terschelling

(7)

7 In this research two blowouts within this area were studied and compared (figure 2). The northeast blowout has been fully stable for at least the past 25 years; it most likely stabilized around 1945. The southwest blowout was still active in 1991 - 25 years ago - and was therefore not reactivated (Kooijman, personal communication, 8-5-16; Van der Meulen et al., 1996). At some point during the last 25 years this blowout stabilized as well. The exact timing will be determined by air photo analysis as part of the larger research effort studying management measures in the Grey dunes (Kooijman, personal communication, 1-6-16).

2. Theoretical framework

Nitrogen emissions into the atmosphere have increased substantially over the 20th century, mainly in the form of nitrogen oxides (NOx) from industrial activities and ammonia (NH3) from agriculture practices (Hicks et al., 2011). The input of nitrogen compounds alters the natural ecological balance of a habitat. This results in an increase in species that prefer high rates of nitrogen supply and has an unfavorable effect on more sensitive species. In addition, there is also evidence for a net loss in biodiversity (Hicks et

al., 2011). Over the last two decades nitrogen emissions have decreased noticeably,

however the Netherlands remains one of the countries where critical loads of nitrogen deposition are still exceeded (Slootweg et al., 2014). It was calculated that 87% of all Natura2000 areas in the Netherlands are at risk for eutrophication (Ibid.). Even though deposition levels have decreased, the effects of nutrient overload are still evident and often long lived (Witz, 2015). Nitrogen emissions have three major impacts on soil and vegetation, namely eutrophication, acidification and direct toxicity (Ibid.). These processes can negatively influence the functioning of a wide range of ecosystem services in the coastal dunes.

The soil at the study site has a low lime content and low upper soil layer pH, resulting in a low to absent calcium carbonate buffer capacity (Kooijman et al., 2005). The availability of N and P is even higher in lime-poor soils than it is in lime-rich soils

Figure 2. GoogleEarth image (2005) of the fieldwork area Eldorado.

Slightly lighter areas indicate blowout activity.

(8)

8 (Kooijman & Besse, 2002). Reactivation exposes the underlying soil layer that has a higher pH (Ibid.). When this newly surfaced sand is then transported by the wind and deposited in the surrounding area, the pH of the soil in that area increases, partly restoring the buffer capacity of the soil. This soil layer can serve as a substrate for new pioneer vegetation (Van der Meulen et al., 1996). A difficulty in achieving this is that the dominant Marram grass (Ammophila arenaria) reestablishes itself quickly after reactivation, through its deep roots (Kooijman et al., 2005). This limits growth of desired pioneer species. Grey dunes are fixed coastal dunes with low aeolian activity as a result of stabilization by vegetation (Provoost et al., 2002). The surface of Grey dunes usually shows a high coverage with mosses and lichens, which accounts for their recognizable Grey color (Witz, 2015). Grey dunes are also called open dune grasslands and form important habitat in the coastal dune landscape, because of their high biodiversity in both plant and animal species (Provoost et al., 2002). They have priority in the habitat directive of the EU and demand special attention for conservation and management as biomass removal seems essential to grassland preservation (Ibid.).

Active blowouts contribute to a dynamic system, making the landscape more resilient. This dynamic activity is generated by wind, which is the most important transport agent in coastal dunes (Witz, 2015). Wind erosion has complex effects on the landscape and creates many different geomorphological landforms. Blowout activation can take place naturally, for instance, by intense storm events or topographic acceleration of the airflow over a dune crest, or by repeated slope wash and subsequent wind erosion, creating a more gradual activation process (Hesp, 2002; Jungerius, 2008). Blowouts can also be reactivated as a management measure, in this case the vegetation and topsoil layer are removed, until a depth of 30 to 50 cm (Van der Meulen et al., 1996).

The shorter the distance to the sea, the more likely it is that there are blowouts, as the wind arrives at the coast in full force. Active blowouts are eroded by the wind at the trailing edge and grains then accumulate at the leading edge (figure 3). This causes the blowouts to slowly migrate in the landscape in an upwind direction (Witz, 2015). After the blowout is initiated the process self-reinforces and blowouts can rapidly expand, but over time the blowout will stabilize naturally (Kooijman et al., 2005).

Figure 3. Upwind migration of blowouts (Jungerius, 2008).

(9)

9

3. Methodology

The workflow gives an overview of the stages within this research (figure 4). At the base of this study is a supporting literature review. The main data collection methods were fieldwork and laboratory analysis of the samples. All the collected data was analyzed using statistics and GIS software. The fieldwork conducted in the research area formed a vital part of this research; it took place over a two-week period during which four transects and a grid were established, encompassing the blowouts and surrounding area (figure 5 & 6). In total there were 38 transect points and 203 grid points. In table 1 below the sample and measurement amounts and types are summarized.

Fieldwork Transect sampling Gridpoint sampling Laboratory analysis Bulk density pH & EC

C & N content of soil

GIS processing Gridpoint analysis (interpolation) Transect analysis (extrapolation) Statistical Analysis Descriptive statistics Pearson correlation coefficients

Two sample T-tests

TRANSECTS GRIDPOINTS

- Coordinates - Soil profile

- Two pF-ring soil samples - Types (%) of vegetation/bare

sand

- Coordinates

- Thickness of O and Ah layer - Types (%) of vegetation/bare sand

Blowout A 19 samples (12 long, 7 short) 91 measurements Blowout B 19 samples (12 long, 7 short) 112 measurements

Total 38 samples 203 measurements

Table 1. Overview of type and number of samples and measurements made during the fieldwork. Figure 4. General overview of workflow

(10)

10 Per blowout – including accumulation zone – two transects were made, one long transect with twelve points and one short transect with seven points (figure 6). At each point a soil sample was taken using a metal pF-ring from the topsoil layer till approximately 5 cm depth. The total soil sample amount in volume is 200 cm3 _{per sampling point. This sample} was used for further analysis of bulk density, pH, EC and C & N content. A detailed description of all measurements is included in table 2 below. Furthermore, surface characteristics were recorded into the categories bare sand and vegetation coverage and soil profiles were made till a depth of 120cm. In addition to the transects a grid was made consisting of 14 rows, the layout of the grid can be seen in figure 5. At the grid points only the thickness of the O & Ah layer and the surface characteristics were described. The approximate interval between the points was 13m between each row and 14m between each column.

Soil Properties Outline of laboratory measurements

Bulk Density

A subsample of 20g was put in the oven at 105°c for 48 hours and cooled down for 15 minutes. The sample was weighed before and after the drying period to calculate the weight loss between the wet and dry sample. Based on the below formula the bulk density was then calculated and converted into m2_.

Total wet weight x Dry subsample

Bulk density = --- Wet subsample weight x Volume original sample

EC & pH

After removing the subsample for bulk density measurement the remaining samples were put in the oven at 70°c for 48 hours. The samples were then sieved with a 2mm sieve. A subsample of 10gr per sample was mixed with 25ml of demineralized water, which was then put on a shaker for two hours, followed by a resting period overnight and finally being put on the shaker for another 20 min the next day. The pH and EC were then measured using an electrode.

C & N content

After taking out the subsample for bulk density, what remained of the samples was put in the oven at 70°c for 48 hours. The samples were then sieved with a 2mm sieve. A small amount from the dried & sieved sample was grounded at a rotational speed of 400rpm in a grinder. Two subsamples of approximately 50mg were analyzed in the CHNS analyzer for carbon and nitrogen content (%). This was then converted into g/m2_{by multiplying it with the bulk} density per m2_.

Table 2. Descriptions of the measurements made in the laboratory for each of the measured characteristics.

(11)

11 The gathered data from the fieldwork and laboratory analysis was digitalized in Microsoft Excel 2011. Statistical analysis of this data was mostly done with MATLAB software. Conversions and simple calculations were performed in Excel. Initially all the descriptive statistics were calculated for the data –mean, min, max, standard deviation – after which two sample T-tests were done for all parameters. This compared blowout A to blowout B and tested for any significant differences between the values. In addition, Pearson correlation coefficients were calculated for a number of parameters. For pH and bare sand Figure 6. Layout of the transect points in the research area of Eldorado (GoogleEarth, 2005). Yellow polygon is blowout A, red polygon is blowout B. Figure 5. Layout of the grid points in the research area of Eldorado (GoogleEarth, 2005). Yellow polygon is blowout A, red polygon is blowout B.

(12)

12 in particular, in the case of a high correlation coefficient a pH map could be produced by extrapolating the data. However, there was no high correlation between the collected data of these parameters.

The grid analysis was done using GoogleEarth and ArcMap 10.2. The coordinates – measured with a Trimble ArcGIS GPS device – were entered in GoogleEarth after which the grid points were joined with data from the examined parameters in ArcMaps. Based on this data, maps were produced for the O layer, Ah layer, bare sand coverage and vegetation coverage. The interpolation method used to generate these maps was the inverse distance weighted technique (IDW). This technique estimates the cell values by averaging the values of the sample data points of each processed cell; the closer a point is to the cell being estimated, the more influence or weight it has in the averaging process (ESRI, 2015).

4. Results

4.1 Soil development

There was a significant difference between both blowouts in terms of soil development. Both the O and Ah layers were thinner in the recently stabilized blowout (B) (table 3; figure 8 & 9), indicating less soil development. All soil profiles could be classified as Arenosols as a result of their high sand content, originating most likely from quartz-rich parent material (IUSS working group, 2014). A few soil profiles in the valleys (AL2,AL3,AL4;BL3) displayed oxidized iron mottling. All these profiles had a high moisture content that further increased with the depth of the soil. At AL3 the groundwater level was reached at 20cm depth. The groundwater level in blowout B was lower and started at 45cm depth. The thickest Ah layers in both blowouts were located in the valley (figure 9), especially where there was willow vegetation. A buried Ah layer was found at AL4 at 1m depth. Likewise, the soil profile at BL2 showed a buried Ah layer starting at 110cm. These earlier Ah layers were likely formed during vegetation succession when the dunes were fore dunes, located closer to the beach zone.

Further towards and within the accumulation zone, the Ah layer thickness decreased (figure 9)- indicating a later start in soil development at these locations. In contrast to the northeast and southwest edges of blowout A that showed a higher percentage of bare sand, both the deflation and accumulation zone had a full vegetation coverage (figure 7). On the northeast side this was most likely caused by grazing animals and hikers. For the southwest side it is suspected that sand was blown over from the adjacent blowout B. The soil profiles in blowout B were in general less developed, with thinner Ah layers for the whole southeast half of the blowout, all part of the deflation zone

A: Stable B: Recently active Significant difference

O layer 1.55 cm 0.29 cm Yes

Ah layer 14.76 cm 4.58 cm Yes

Bare sand 4.62 % 19.1 % Yes

Bulk density 1.05 g/cm3 _{1.21 g/cm}3 _Yes

(13)

13 (figure 9). At BL11 and BK18 the Ah layer was in fact completely absent. On an air photo from GoogleEarth (2005) it seemed that there was still some blowout activity in the deflation zone and bare sand in the accumulation zone of blowout B, consequently soil development in this area only started recently. Especially the whole east corner of the blowout itself had remained active for a long time and still showed some activity. The vegetation in this area was sparse, mainly consisting of Viola curtisii and Ammophilia

arenaria. The bare sand distribution map depicts that this area had a high bare sand

coverage (figure 7) and similarly thin Ah layers (figure 9).

Bulk density

The bulk density of the soil is an indicator of soil compaction, which affects properties such as water infiltration, soil porosity and plant nutrient availability (Chaudhari et al., 2013). As a result, the bulk density is a good indicator for soil quality and length of soil development. The

bulk density in blowout A ranged between 0.37 g/cm3_{and 1.36 g/cm}3_{and has a strong} negative correlation with the amount of OM, meaning that the bulk density usually decreases as the OM content increases (Chaudhari et al., 2013). The two lowest bulk density measurements in blowout A both came from locations with a relatively thick O and Ah layer. At AL1 there was a thick moss layer present and at AL3 there was a peat like soil, both high in OM content, which is lighter than sand. The peat formation in the valley combined with the high water level indicates poor drainage. The excess water retards decomposition and results in high levels of soil development because of rapid accumulation of OM (Sevink, 1991). Based on this and the relation between bulk density and OM, it was clear that there was a high SOM content at these locations. The bulk density of the samples in blowout B ranges between 0.61 g/cm3_{and 1.51 g/cm}3_{. Overall,} blowout B had a higher range and mean bulk density (table 3). Sand content and bulk density have a high positive correlation (Chaudhari et al., 2013). Based on the higher average bulk density in blowout B, it is likely that soil there has a higher sand content. Besides this the correlation between bare sand and bulk density turned out to be significant (table 4), with bulk density becoming higher as there is more bare sand.

Summary

Based on the findings blowout A likely has a higher nutrient availability and a higher water holding capacity, both because of a higher OM content. Moreover, this blowout overall has a thicker O and Ah layer, indicating a higher degree of soil development. In conclusion, blowout activity seems to result in a lower degree of soil development and affects the development and ecosystem even after stabilization. The effects of active blowout processes are still noticeable in the soil and there is a significant difference between the blowout that stabilized approximately 70 years ago and the blowout that stabilized in the last 25 years.

Parameters Bulk density

Bare sand 0.36

O layer -0.57

Ah layer -0.36

(14)

14

Figure 7. Bare sand (%) distribution map based on grid data.

Figure 8. Distribution map of the thickness of the O layer (cm) based on grid data.

Figure 9. Distribution map of the thickness of the Ah layer (cm) based on grid data.

(15)

15

4.2 pH dynamics and Electrical Conductivity

In blowout A, the pH ranges from 4.53 to 5.97. The pH range in blowout B is slightly lower -4.23 to 5.77- but there is no significant difference. The measured pH levels are in line with the known lime-poor soil in the area, however usually soils without an accumulation layer have a pH of around 4 (Van der Meulen et al., 1996). In this case both the pH in the stable and recently stabilized blowout are slightly higher. Still, the pH in blowout B was lower than expected, since pH is expected to increase as a result of blowout activity. Thus the pH in blowout B was expected to be higher than the pH in blowout A. Both soils have a pH of around 5 and can therefore be classified as acidic soils (Christopherson, 2013). This means that acidification of the soil was not effectively counteracted in the recently stabilized blowout.

The pH measurements did not have a high correlation with bare sand coverage (Appendix C, figure 1; table 4), which was initially expected. Aside from the C content and C/N ratio, there were no other significant correlations between pH and the other measured parameters (table 4). As a result of a lack of correlation between pH and any of the grid parameters, no pH map could be produced. The negative correlation between carbon and pH (table 4) can

be explained by the fact that the pH is usually higher in recently exposed sand. This sand does not have a high OM content and thus has a low C content. Over time the OM content increases, because of vegetation growth and the pH decreases as a result of leaching.

pH 5.29 5.12 No EC 105.1 mS/cm 80.2 mS/cm No Parameters pH EC Bare sand 0.12 -0.30 Bulk density 0.15 -0.79 C/N ratio -0.48 0.51 C content per m2 _-0.34 _0.66 N content per m2 _-0.12 _0.56 4 4.5 5 5.5 6 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 pH le ve l pH level A: Fully stabilized blowout B: Recently stabilized blowout Table 3. Means of pH and EC and T-test outcomes.

Table 4. Pearson Correlation Coefficients

(16)

16

Electrical Conductivity

The EC of a soil varies depending on the amount of moisture held by soil particles (Grisso

et al., 2009). Consequently, soils with a high amount of sand have a low conductivity. The

lowest EC in blowout A was measured at the northeast edge of the blowout (AK13) where there was 50% bare sand coverage. The highest EC was measured at AL3 in peat soil. This corresponds to the expectation that the EC changes proportionally to the water holding capacity of the soil (Grisso et al., 2009). Overall the EC measurements are lower in blowout B than in blowout A, including a lower EC range and mean (figure 11; Appendix B). This is in line with the thinner O and Ah layers in the recently active blowout. The blowout stabilized later and there was a shorter period of soil development, resulting in less SOM and a higher share of sand.

The sampling points where the EC was low are also expected to have a low porosity volume, because electricity is transmitted easier in proportion to the porosity (Grisso et al., 2009). This means that both blowouts have a higher porosity in the valleys and thus a higher water holding capacity. Consequently, the potential vegetation

cover and biomass is higher. This is confirmed by both the average vegetation cover and willow and shrub vegetation in the valleys (Boersma, 2016). Earlier research found a significant negative correlation between bulk density and EC (Chaudhari et al., 2013). Bulk density values are usually high when the soil has a high sand content, the EC on the other hand decreases with increased sand content. The negative correlation held true for the collected data in this research with a correlation coefficient of -0.79 between EC and bulk density.

Summary

Based on these results it cannot be concluded that blowout activity counteracts acidification on the long-term in lime-poor soils, despite the findings of earlier studies - that blowout activity counteracts acidification in the short term (Van der Meulen et al., 1996; Kooijman et al., 2005). With regard to EC there is also no significant difference between the blowouts. The EC measurements do, however, indicate that there is a higher porosity and water holding capacity in the valley and that there is a higher sand content in the soil for the rest of the blowout, especially in the accumulation zone (figure 6, points 10/11/12).

Figure 11. Measured electrical conductivities along transects 0 100 200 300 400 500 600 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 mS /c m Electrical conductivity A: Fully stabilized blowout B: Recently stabilized blowout

(17)

17

4.3 Nitrogen and Carbon dynamics

The average N content of the soil is lower in blowout B than in blowout A (table 5). This can be explained by the lower vegetation cover and higher bare sand cover in blowout B (figure 7), since N availability is directly dependent on the decomposition of OM and thus biomass input (Kooijman & Besse,

2002). Because of this the N content shows a strong correlation to the C content. The N dynamics are similar to the bulk density values in the way that low bulk density values generally have a high OM content and thus a comparably higher N availability (Chaudhari et al., 2013). However, N can only be taken up by the plant after mineralization. In both blowouts the highest nitrogen contents per m2 _{were found in the} valley and the lowest near the high edges (figure 13). The N content showed a high correlation with the O layer thickness, explaining around 30% of the variance in this parameter. By fitting a trend line through the data (Appendix C, figure 2), a function could be derived to calculate the N content for all the grid points. With this data the N content distribution map could be generated (figure 14). The finding that the N content per m2_{is significantly lower in blowout B is important, because it indicates that} eutrophication of the soil is indeed lower in the recently active blowout. The lower N content of the soil results in less encroachment with Ammophilia arenaria, which is one of the goals of reactivation. Though, at the moment encroachment is already limited by livestock grazing in the area.

Biomass input is directly related to the C content, since the C content of vegetation is approximately half the oven dry weight (Magnussen & Reed, 2005). Locations with a high content of OM are thus expected to have a high soil C content. Conversely, locations

Soil N content 89.9 gr/m2 _{62.6 gr/m}2 _Yes Soil C content 1083.6 gr/m2 _{775.3 gr/m}2 _No

Soil C/N ratio 12.4 11.12 No

Parameters C/N ratio C content (m2₎ _{N content (m}2₎

Bulk density -0.80 -0.80 -0.65 O layer 0.43 0.59 0.59 Ah 0.36 0.42 0.37 Bare sand -0.52 -0.40 -0.36 0 50 100 150 200 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 g/ m 2 Nitrogen content A: Fully stabilized blowout B: Recently stabilized blowout Table 5. Means of C & N and T-test outcomes.

Table 6. Pearson Correlation Coefficients

(18)

18 with a high percentage of bare sand are expected to have a lower C content. A difference between blowouts would therefore be expected, however the C content in blowout B is not significantly lower than in blowout A (table 5). With a 0.1 significance level on the other hand, there is a significant difference (Appendix B). The lower average C in blowout B indicates that there has been a larger amount of OM input in blowout A over time (table 5). Meaning that, because blowout A stabilized earlier, the total OM input there has been higher. The highest C content in blowout A was measured in the peat soil (AL3); the lowest at the southwestern edge (AL19) (figure 15). The latter sample point might have been influenced by sand blown over from the adjacent blowout. Most of the vegetation on the southwestern edge consisted of moss and grass, classified as vegetation type two, characteristic to a later

successional stage (Appendix D). The highest C content in blowout B was found at BL3, where the vegetation consisted of willow and shrubs (figure 15). The lowest C content (BL11) was measured at a location that had a higher bare sand than vegetation coverage (figure 15).

The mean C/N ratio is lower in blowout B, although not significantly. The lower ratio in B implies that there is more N available in the soil, but this is relative to the SOC content. The total amount of N is still lower in blowout B, indicating that there is less eutrophication occurring than in blowout A.

Figure 14. Nitrogen content per m2_{distribution map based on transect data, extrapolated to grid points.}

0.00 500.00 1000.00 1500.00 2000.00 2500.00 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 gr /m 2 Carbon content A: Fully stabilized blowout B: Recently stabilized blowout

(19)

19

Summary

To conclude, blowout activity significantly reduced the N content per m2_{, even in the} long-term, resulting in less eutrophication of the soil compared to the earlier-stabilized blowout. Both the C content and C/N ratio are not significantly different, but do have lower means for blowout B. The lower average C content in B does indicate that blowout activity could have resulted in less vegetation growth, but because of the lack of a statistically significant difference, this cannot be stated conclusively.

5. Discussion

5.1 Influence of blowout activity on soil development

Soil patterns in dunes are typically complex, as a result of the interaction and alternation between geomorphological and biological processes (Sevink, 1991). Soil development is highest where geomorphological activity is low and biological processes are dominant (Van Bolhuis, 1995). This is the case in blowout A and the northwestern half of blowout B. The nutrient content in the soil is higher at these locations than at places with a deep accumulation layer, where the nutrient-poor sand surfaces (Ibid.). A thin accumulation layer however, temporarily increases the nutrient availability by re-calcifying the soil (Ibid.). The clear negative correlation between bare sand and C and N content in this study indicates that bare sand areas have a lower nutrient content. None of these bare sand areas have a thin accumulation layer, so this is in line with the findings by Van Bolhuis (1995).

A study by Jones et al. (2008) concluded that soil development in sand dunes follows a sigmoid curve, with a slow initial rate, increasing rapidly as time progresses and then stabilizing again. At the start a lack of soil moisture and nutrients limits the initial vegetation succession. This pattern accounts for the bare sand in the east corner of blowout B, which has remained bare since at least 2005, based on GoogleEarth (2005) imagery. The low biomass production there strongly retards soil development (Provoost

et al., 2002). It indicates that by limiting the biomass input through management

measures soil development can be inhibited in stabilized dune areas (Jones et al., 2008). While seemingly logical, this is important for dune management. After initial succession, the input of organic matter improves moisture retention and nutrient availability, positively reinforcing the soil development rate (Provoost et al., 2002).

For Grey dunes soil formation and pluvial processes are dominant over aeolian processes (Jungerius, 2008), though blowouts are formed by aeolian processes. The most influential wind is the Northwesterly wind, combined with the West-Southwesterly wind that is usually the strongest along the Dutch coast (Van der Meulen et al., 1996). Both winds explain the prolonged activity on the eastern slope of the recently stabilized blowout B. Slope wash enables blowout formation by eroding the top layer (Jungerius, 2008). This explains why the area that has remained active for the longest is a slope, where these processes have significant influence. Bare zones are water repellent and especially after a long dry period the sand is remarkably impermeable (Jungerius, 2008). Soils can dry out to 20 cm deep and can heat up to 60°C when exposed to the sun (Provoost

(20)

20

et al., 2002). In Grey dunes these conditions result in slope wash, with the water

accumulating in the valley (Jungerius, 2008). The water infiltrates at the lowest point and at places where it is retained by vegetation. This pattern is noticeable in both blowouts, with high soil moisture contents and high water levels, caused by poor drainage, in the valleys.

5.2 Counteracting acidification

The findings of this research indicate that in the long-term acidification is not counteracted within naturally active blowouts. However, short-term monitoring indicated a significant increase in pH values to between 7 and 8 in accumulation zones within the same area (Van der Meulen et al., 1996; Van Bolhuis, 1995). Moreover, the reactivated blowout studied by Van Bentum (2016) and Van den Berg (2016) did show a significant difference, with a higher pH in the reactivated blowout. The leaching of lime from the recently exposed sand might explain why there was no effect on acidification in the studied blowouts. Dune soils are prone to leaching as a result of a relatively low water storage capacity (Sevink, 1991). In temperate climates rates of leaching and acidification are especially high, because of a precipitation surplus and low decomposition rate (Sevink, 1991). The sand that surfaces when the blowout is active has a higher lime content. These bare areas have a higher pH, but the findings did not indicate any correlation between these parameters. Since the blowout was already active before 1991 (Van der Meulen et

al., 1996) and most of the blowout has been stable for at least a number of years, leaching

most likely reduced the pH again. This does not exclude that on the short term acidification was in part counteracted, however on the long-term the soil re-acidifies.

In general, soils with a low pH (4-5) have a low decomposition and mineralization rate, which increases the humus content and soil development (Provoost et al., 2002; Jones et al., 2008; Provoost et al., 2011). Moreover, the humic acids released during soil development further enforce leaching and decrease the pH, forming a positive feedback loop (Provoost et al., 2002). It would therefore be expected that the earlier-stabilized blowout, with a significantly thicker O layer and lower bulk density, also has a significantly lower pH. However, this was not the case; the processes might still have had an effect, but the resulting difference between the blowouts is not significant. The pH measurements might have been higher and significantly different between the blowouts, if samples had been taken in the east corner of blowout B as well, which still shows slight blowout activity. In this area there was a relatively high amount of Viola curtisii vegetation, a plant that grows well in lime-rich soils (Dijkstra, 2001). However, the lack of correlation between pH and bare sand for both blowouts contradicts this idea. Still, it is important to take into account that the outcome was also influenced by the location of the transect points. At locations with 50% bare sand (AK13; BL12) the pH was very similar, 5.29 and 5.26 respectively. Even the highest pH in either blowout had a similar bare sand coverage (AK19: 10%: 5.97; BK19: 5%: 5.77). Surprisingly, both maximum pH measurements had the same transect location. This could indicate that the pH levels follow a similar distribution throughout the blowout (in line with the lack of significant difference), which supports the idea that blowout B has returned to its former pH levels after stabilization.

(21)

21

5.3 Nutrient dynamics

Increased atmospheric N deposition has been shown to increase soil development (Jones

et al., 2008). It increases plant productivity and OM input, which boosts soil fertility and

nutrient retention (Provoost et al., 2011). Earlier research also indicated that N deposition can alter soil processes (Ibid.). Nonetheless, Van Boxel et al., (1997) noted that over a course of three years (1991-1994) reactivated blowouts remained active -and even grew- despite increased inputs of N as a result of atmospheric deposition. In the longer term, N deposition might still play an important role in stabilization, especially for naturally active blowouts. With regard to this study, the higher N content in blowout A is in line with the higher level of soil development and higher vegetation cover.

In the Waddendistrict most of the mineralized N is transformed to nitrate (Kooijman & Besse, 2002). If the pH of blowout B has decreased as a result of leaching, then part of the N content, in the soluble form of nitrate, in the soil might also have been leached. Sandy and coarse textured soils have a low water holding capacity and because of this have a greater potential for leaching (Lamb et al., 2014). While this might play a role, the positive feedback loop of N deposition – increasing plant productivity, resulting in increased OM input, further increasing N content- is expected to play a larger role. Therefore, for the recently stabilized blowout, the N content is likely lower due to a lower total OM input, caused by both a later stabilization and a lower vegetation coverage. Overall, the nutrient availability for the Waddendistrict is high (Kooijman & Besse, 2002). The mean pH levels in both blowouts seem to indicate that the nutrient P does not play a limiting factor in plant productivity (figure 16). The optimum pH level for P availability was found to be approximately 5.3 (Kooijman & Besse, 2002). For lower pH levels P is immobilized in Al and Fe compounds (Provoost et al., 2011). Usually, the availability of P is one of the main limiting factors in plant productivity in Grey dunes, however for the Waddendistrict the P availability is relatively high and N is the limiting factor (Kooijman & Besse, 2002). This would indicate that Eldorado is even more sensitive to increased atmospheric N deposition.

The C contents in the valley soils were much higher than the rest of the values, as a result of the peat like soils at these locations. In consequence, these had a large influence on the outcome when testing for a significant difference. When the valley soils are not taken into account (AL3,AL4;BL3,BL4) there is a significant difference, with a probability value of 0.0232 (Appendix B). The excluded outliers had a large amount of OM, resulting in a high C content, which decreased the difference between the blowouts. When not included in the T-test, the outcome is in line with the significantly different O layer, bulk density and the higher amount of bare sand in blowout B. From this finding it can be concluded that blowout activity did result in lagged vegetation growth. The first species to establish during plant succession are commonly algae (van Bolhuis, 1995). Algae and micro-organisms present in the soil during the initial soil development stages lower the C/N ratio of the soil (Kooijman et al., 2005). This could explain why the less developed soil in the recently stabilized blowout has a lower C/N ratio. The ratio is also lower because there is less OM and thus less C in the soil. If taking into account the sigmoid curve of soil development (Jones et al., 2008), it could be that changes in N content have a larger effect on the soil and vegetation in the recently stabilized blowout, relative to the early stabilized blowout, which already has a high N content and high vegetation coverage.

(22)

22

5.4 Differences between the natural reference and reactivated blowout

The blowouts studied by Van Bentum (2016) and Van den Berg (2016) did show a significant difference in pH, however, there is no significant difference between the reactivated blowout and the recently stabilized blowout (Van Bentum, 2016; Van den Berg, 2016; figure 15). It could be that the effect of reactivation on increasing the pH is more significant than the increase caused by naturally activated blowout processes. Reactivation of a blowout takes place by removing the vegetation and topsoil layer (Van der Meulen et al., 1996). The topsoil layer contains a high amount of soil organic matter and is therefore less easily eroded by the wind (Kooijman et al., 2005). The new topsoil has a much higher pH and lower nutrient availability. Natural activation, on the other hand, is often more gradual and exposes less of the underlying soil. This explanation however would suggest a significantly higher pH in the reactivated blowout, which is not the case. On the other hand, both pH levels are relatively high for the area, suggesting that blowout activity did affect the pH. Usually, soils without an accumulation

layer have a pH of around 4 (Van der Meulen et al., 1996). Because the pH of the recently stabilized blowout was not measured 25 years ago (Van der Meulen et al., 1996), it is hard to say if there has been a decrease or if pH levels have remained stable during blowout stabilization. Overall, the high plant biodiversity in blowout B seems to indicate that acidification has not negatively affected the habitat.

The differences in N content of the two natural reference situations seem to be smaller than when comparing a reactivated blowout to a stable one (figure 17). Removal of the topsoil and vegetation lowers the N content of the soil, because the biomass input into the soil, and thus a share of N, is removed from the system (Lamb et al., 2014). This makes the N lowering effect of reactivation more significant. Within naturally active blowouts, none of the N is actually removed; instead the process triggering blowout development only moves OM of both vegetation and soil. It is likely that part of the OM and thus N still stays in the soil.

1084 90 775 63 1028 56 620 39 0 200 400 600 800 1000 1200

C content N content C content N content

C & N content

Stable Recently stabilized Non-reactivated Reactivated

Figure 16. pH levels of the blowouts in this study (Stable & Recently stabilized) compared to those studied by Van Bentum & Van den Berg (2016).

Figure 17. C and N contents of the blowouts in this study (Stable & Recently stabilized) compared to those studied by Van Bentum & Van den Berg (2016).

5.29 5.12 4.51 5.09 4 4.2 4.4 4.6 4.8 5 5.2 5.4

Stable Recently Stabilized Non-reactivated Reactivated

(23)

23 Moreover, blowout activation of the recently stabilized blowout started earlier than for the reactivated blowout. Since the recently stabilized blowout was already fully active before 1991, while the reactivated blowout was initialized during that year (Van der Meulen et al., 1996; Kooijman, personal communication, 8-5-16). As a result, the rebuild up of N content in the soil through atmospheric N deposition started earlier in the recently stabilized blowout, explaining its higher N content.

5.5 Influence of additional management measures on the area

Since the initial research took place the area has undergone significant changes. For a number of years, the area has been populated by a group of Galloway cows and Exmoor ponies (Lamers, 2011). These animals have limited the encroachment of Marram significantly and have overall improved biodiversity in the whole area. The grazers were so successful that their numbers had to be reduced because there was no longer enough grass available (Lamers, 2015). The impact of grazing on soil stability is especially high for soils susceptible to accelerated erosion, based on the spatial density of dune crests (Blanco et al., 2008). Livestock grazing can generate and accelerate soil erosion processes when used intensively (Blanco et al., 2008). Van Bolhuis (1995) found that because of a lack of an iron-oxide coating of the sand grains the sand in the Waddendistrict is more easily eroded by water and wind. For Eldorado, the initialization of soil erosion is desirable, because it helps activate blowouts and aeolian activity.

In combination with the management measure of livestock grazing, managers created a semi-natural water habitat to attract birds and other wildlife. This area is located behind the studied blowouts and part of the dunes was excavated in order to realize the plan. This likely affected the accumulation zone of blowout B was, which might have influenced the grid point data from the last row on the southeast side. According to Jungerius (2008) these management measures to ‘restore nature’ and increase biodiversity form a real danger, by destroying often centuries old geomorphic systems. On the other hand, the measure seems to have been successful in increasing faunal biodiversity.

5.6 Management recommendations

While source-directed policies aimed at reducing nitrogen dioxide emissions have led to a small decrease in N deposition, the deposition levels are still high above natural reference level (Provoost et al., 2011). Symptom mitigation is therefore necessary to maintain the coastal dune habitat. The findings of this research indicate that reactivation of blowouts would contribute to reducing eutrophication and encroachment and that it might reduce acidification, despite inconclusive results for the latter. This management strategy can be used in combination with other measures, such as livestock grazing. The removal of biomass and incited erosion seem to have had positive effect on the Eldorado area. As noted by Provoost et al. (2002), biomass removal is essential for preservation of the Dune grassland. The effects of grazing management on soil development are however complex and remain a subject for further research (Ibid.).

(24)

24

6. Conclusion

This study assessed the long-term impact of active blowout processes on soil development and acidification and eutrophication of the soil by comparing a recently stabilized to a stable blowout. It can be concluded that active blowout processes limit soil development as a result of aeolian dynamics and that the effects of these processes are noticeable even in the long-term. Soil development was higher at locations where there was a higher soil moisture content and poor drainage, which limited the decomposition rate. Overall, the soil in the earlier-stabilized blowout was more developed with significantly thicker O and Ah layers. This was due to a higher biomass input to the soil and a longer development period.

Furthermore, it cannot be concluded that active blowout processes counteract acidification in lime-poor soils on the long-term based on the current findings. There were no significant differences in pH between the stable and recently active blowout. While the soils were acidic, the pH was relatively high for the lime-poor sand in the area. It is plausible that the pH increased shortly after activation of the blowout, but decreased again over time. The N content was significantly lower in the recently stabilized blowout. This indicates that active blowout processes counteract eutrophication, even over a longer term. The lower N content also explains the lower degree of soil development in the recently stabilized blowout and vice versa. The higher N content in the early-stabilized blowout contributes to increased soil development, by increasing the biomass input through a positive feedback loop. Besides this, the significantly different C content, when the valley soils are excluded, supports the idea that vegetation growth initially lags as a result of blowout activity.

Overall, the hypothesis can be partly accepted; naturally active blowout processes have resulted in a lower degree of soil development in the recently stabilized blowout and have counteracted eutrophication. Whether acidification has been counteracted cannot be determined conclusively, because there is no significant difference in pH. While this study assessed the implications of naturally active blowouts on soil quality, it does not directly cover the effects of reactivated blowouts. Hence, it should be kept in mind that there could be unaccounted differences between naturally active and reactivated blowouts. In conclusion, the net effect of active blowout processes seems to have been positive on both the soil and plant biodiversity. Therefore, reactivation appears to be a good habitat restoration strategy.

(25)

25

References

Blanco, P. D., Rostagno, C. M., Del Valle, H. F., Beeskow, A. M., & Wiegand, T. (2008). Grazing impacts in vegetated dune fields: predictions from spatial pattern analysis. Rangeland ecology & management, 61(2), 194-203.

Boersma, M. (2016). The long term impact of active blowouts on vegetation

characteristics in the Grey dunes of Eldorado, Terschelling. Bachelor Thesis, University of Amsterdam

Chaudhari, P. R., Ahire, D. V., Ahire, V. D., Chkravarty, M., & Maity, S. (2013). Soil bulk density as related to soil texture, organic matter content and available total

nutrients of Coimbatore soil. International Journal of Scientific and Research

Publications, 3(2), 1-8.

Christopherson, R. W. (2013). Chapter 14 The geography of Soils. Elemental geosystems. Seventh Edition, Pearson Education, Inc.

Dijkstra, K. M. (2001). Duinviooltje, Viola curtisii. Wilde planten in Nederland en Belgie. Retrieved from wilde-planten.nl/duinviooltje.htm

Doekes, E., Nijboer, M. & Bekker, L. (2015). Deel I Plan-MER over het programma aanpak stikstof 2015-2021: Een integraal milieukundig onderzoek naar de effecten van de Programma Aanpak Stikstof (PAS). Ministerie van Economische Zaken en Ministerie van Infrastructuur & Milieu.

Environmental systems research institute - ESRI (2015). Comparing Interpolation Methods, ArcGIS Pro. Retrieved from http://pro.arcgis.com/en/pro-app/tool-reference/3d-analyst/comparing-interpolation-methods.htm

Google earth (2005). Satellite imagery 31-12-3005. 530_{23’47.90’’ 5}0_{14’30.58’’} Eldorado, Terschelling. Aerodata International Surveys.

Grisso, R. D., Alley, M. M., Holshouser, D. L., & Thomason, W. E. (2009). Precision Farming Tools. Soil Electrical Conductivity. College of Agriculture and life sciences, Virginia polytechnic institute and state university.

Hesp, P. (2002). Foredunes and blowouts: initiation, geomorphology and dynamics.

Geomorphology, 48(1), 245-268.

Hicks, W. K., Whitfield, C.P., Bealey, W.J., Sutton, M.A. (2011). Nitrogen Deposition and Natura 2000: Science and Practice in Determining Environmental Impacts.

COST729/ Nine/ESF/CCW/JNCC/SEI Workshop Proceedings, published by COST. Available at: http:// cost729.ceh.ac.uk/n2kworkshop

IUSS working group - WRB (2014). World reference base for soil resources, International soil classification system for naming soils and creating legends for soil maps. World Soil Resources Reports No. 106. FAO, Rome.

Jones, M. L., Sowerby, A., Williams, D. L., & Jones, R. E. (2008). Factors controlling soil development in sand dunes: evidence from a coastal dune soil chronosequence. Plant

(26)

26 Jungerius, P. D. (2008). Dune development and management, geomorphological and soil processes, responses to sea level rise and climate change. Baltica, 21(1-2), 13-23.

Kadaster/ Topografische Dienst (2015). Topografische kaart van Nederland 1:25.000: 5A Terschelling. Apeldoorn.

Kooijman, A., & Besse, M. (2002). The higher availability of N and P in lime-poor than in lime-rich coastal dunes in the Netherlands. Journal of Ecology, 90(2), 394-403.

Kooijman, A., Besse, M., Haak, R., Van Boxel, J., Esselink, H., ten Haaf, C., et al. (2005). Effectgerichte maatregelen tegen verzuring en eutrofiering in open droge duinen. Eindrapport Fase, 2.

Łabuz, T. (2015). Coastal Dunes: Changes of Their Perception and Environmental Management. In Environmental Management and Governance (pp. 323-410). Springer International Publishing.

Lamers, J. (2011). Grasmaaiers en bosmaaiers. Staatsbosbeheer Terschelling. Retrieved from https://staatsbosbeheerterschelling.wordpress.com/2011/06/14/ grasmaaiers-en-bosmaaiers/

Lamers, J. (2015). Minder gras, dus minder grazers. Staatsbosbeheer Terschelling. Retrieved from https://staatsbosbeheerterschelling.wordpress.com/2015/01/29/minder-gras-dus-minder-grazers/

Lamb, J. A., Fernandez, F. G. & Kaiser, D. E. (2014). Nutrient management: Understanding nitrogen in soils. University of Minnesota.

Magnussen, S. & Reed, D. (2005). Chapter 5 Carbon content estimation. Modeling for estimation and monitoring – knowledge reference for national forest assessments. Food and Agriculture organization of the United Nations – FAO

Provoost, S., Ampe, C., Bonte, D., Cosyns, E., & Hoffmann, M. (2002). Ecology, management and monitoring of dune grasslands in flanders, belgium. Littoral 2002. The Changing Coast, 11-22.

Provoost, S., Jones, M., & Edmondson, S. (2011). Changes in landscape and vegetation of coastal dunes in northwest Europe: A review. Journal of Coastal

Conservation, 15(1), 207-226.

Sevink, J. (1991). Soil development in the coastal dunes and its relation to climate.

Landscape Ecology, 6(1-2), 49-56.

Slootweg, J., Posch, M., Hettelingh, J., & Mathijssen, L. (2014). Modelling and mapping the impacts of atmospheric deposition on plant species diversity in Europe - CCE status report 2014 No. 2014- 0075). Coordination Centre for Effects.

Van Bentum, D. A. (2016). The potential of reactivating blowouts for soil quality in the coastal Grey dunes of Eldorado, Terschelling. Bachelor Thesis, University of Amsterdam

(27)

27 Van Bolhuis, A. (1995). Effect van wind op de dikte en grootte van accumulatiezones Eldorado Terschelling. Universiteit van Amsterdam, Fysisch Geografisch Bodemkundig Laboratorium.

Van Boxel, J. H., Jungerius, P. D., Kieffer, N., & Hampele, N. (1997). Ecological effects of reactivation of artificially stabilized blowouts in coastal dunes. Journal of

coastal conservation, 3(1), 57-62.

Van den Berg, M. (2016). The long term effect of blowout reactivation on the vegetation biodiversity in the Grey dunes of Eldorado, Terschelling. Bachelor Thesis, University of Amsterdam

Van der Meulen F., Kooijman A. M., Veer M. A. C. & Van Boxel J. H., (1996). Effectgerichte maatregelen tegen verzuring en eutrofiering in open droge duinen. Fysisch Geografisch en Bodemkundig Laboratorium, Universiteit van Amsterdam, 232 pp.

Veer, M. & Kooijman, A. (1997). Effects of grass-encroachment on vegetation and soil in dutch dry dune grasslands. Plant and Soil, 192(1), 119-128.

Witz L. (2015). The potential of small-scale blowout activity for landscape diversity in the Dutch Grey dunes. Master Thesis, University of Amsterdam.

(28)

28

Appendix A Transect Data

Label code pH EC Bulk density (g/cm3₎ N (g/m2₎ _{C (g/m}2₎ _{C/N ratio} AL1 4,68 101,4 0,63 51,62 781,95 15,14 AL2 5,17 69,8 1,05 127,65 1725,35 13,42 AL3 5,50 552,0 0,37 151,46 2475,57 16,35 AL4 5,65 158,4 1,01 206,17 1097,70 12,32 AL5 5,69 69,2 1,08 85,72 989,44 11,55 AL6 5,67 111,2 1,16 75,54 807,23 10,70 AL7 5,49 83,8 1,12 54,98 663,00 12,08 AL8 5,72 30,6 1,28 44,57 453,69 10,15 AL9 4,91 129,8 1,04 124,48 1759,73 14,15 AL10 4,73 57,1 1,07 90,36 1175,15 13,06 AL11 4,82 40,6 1,04 127,82 1739,10 13,59 AL12 5,35 148,4 0,98 117,93 1404,37 11,92 AK13 5,29 28,1 1,36 40,56 414,83 10,29 AK14 4,63 71,6 1,00 73,98 1031,47 13,98 AK15 4,53 104,0 1,04 77,97 1217,61 15,60 AK16 5,13 30,2 1,19 46,00 517,35 11,33 AK17 5,61 89,8 1,14 104,54 1276,17 12,23 AK18 5,93 63,6 1,06 75,19 816,52 10,90 AK19 5,97 57,7 1,35 32,25 242,15 7,52 BL1 4,23 65,7 1,23 103,85 1549,41 14,95 BL2 4,48 82,1 1,24 80,79 1028,91 12,80 BL3 4,81 302,0 0,61 134,67 2116,71 15,72 BL4 5,01 121,3 0,89 113,16 1737,85 15,36 BL5 5,07 44,3 1,29 51,95 617,01 11,84 BL6 5,17 112,5 1,51 22,30 173,27 7,84 BL7 5,13 26,3 1,46 36,25 277,70 7,68 BL8 5,11 33,4 1,42 39,49 403,76 10,41 BL9 5,75 84,2 1,29 34,76 284,54 8,22 BL10 5,23 32,2 1,37 34,26 332,02 9,71 BL11 5,10 24,7 1,49 30,03 97,10 3,22 BL12 5,26 65,4 1,06 79,93 978,27 12,26 BK13 4,95 38,8 1,29 61,92 655,37 10,62 BK14 5,52 115,9 1,07 53,42 620,91 11,65 BK15 5,14 90,5 1,14 77,07 1033,11 13,43 BK16 4,96 142,4 0,84 73,63 1219,51 16,57 BK17 5,22 64,4 1,14 64,12 714,97 11,15 BK18 5,40 32,7 1,39 42,32 428,79 10,17 BK19 5,77 45,3 1,26 55,29 461,53 8,54

(29)

29

Appendix B Statistical outcomes

Descriptive statistics*

BLOWOUT A

Parameters Mean SD Min Max

O layer (cm) 1.55 2.30 0 10

Ah layer (cm) 14.76 5.64 6 25

Bare sand (%) 5.26 12.52 0 50

Bare sand in grid (%) 4.62 11.93 0 80

Bulk density (g/cm3₎ _1.05 _0.23 _0.37 _1.36 pH 5.29 0.46 4.53 5.97 EC (mS/cm) 105.12 114.74 28.1 552 Soil N content (g/m2₎ _89.94 _44.49 _32.25 _206.17 Soil C content (g/m2₎ _1083.60 _562.86 _242.15 _2475.57 Soil C/N ratio 12.43 2.14 7.52 16.35 BLOWOUT B

Parameters Mean SD Min Max

O layer (cm) 0.29 0.51 0 2

Ah layer (cm) 4.58 3.57 0 12

Bare sand (%) 10.53 16.82 0 60

Bare sand in grid (%) 19.1 28.27 0 100

Bulk density (g/cm3₎ _1.21 _0.24 _0.61 _1.51 pH 5.12 0.37 4.23 5.77 EC (mS/cm) 80.22 64.48 24.7 302 Soil N content (g/m2₎ _62.59 _30.34 _22.3 _134.67 Soil C content (g/m2₎ _775.30 _559.03 _97.1 _2116.71 Soil C/N ratio 11.16 3.31 3.22 16.57

T-test outcomes*

Parameters Significant difference P value

O layer YES 0,0056

Ah layer YES 9,515467*10-8

Bare sand NO 0,2813

Bare sand in grid YES 2,926837*10-6

Bulk density YES 0,0413

pH NO 0,2267

EC NO 0,4149

Soil N content YES 0,0333

Soil C content NO 0,0989

Soil C content Excluding AL3, AL4 & BL3, BL4 (Valley soils)

YES 0,0232

Soil C/N ratio NO 0,1687

(30)

30

Pearson correlation Coefficients

*Not significant at a 0.05 significance level

Appendix C Correlation graphs

y = 20.581x + 60.015 R² = 0.34866 0.00 50.00 100.00 150.00 200.00 250.00 0 1 2 3 4 5 6

Nitrogen content & O layer

Parameters O

(cm) Ah (cm) Bare sand Bulk density pH EC C/N ratio C content

m2 N content m2 O layer Ah layer 0.51 Bare sand -0.22 -0.50 Bulk density -0.57 -0.36 0.36 pH 0.09* 0.07* 0.12* 0.15* EC 0.65 0.23* -0.30* -0.79 0.01* C/N ratio 0.43 0.36 -0.52 -0.80 -0.48 0.51 C content per m2 0.59 0.42 -0.40 -0.80 -0.34 0.66 0.65 N content per m2 0.59 0.37 -0.36 -0.65 -0.12* 0.56 0.61 0.96

Figure 1. Correlation plot of O layer and N content data. The data shows significant correlation. By fitting a trend line the data can be extrapolated to the grid points, using the function derived from the trend line.

(31)

31

Appendix D Vegetation classification

y = 0.0033x + 5.1787 R² = 0.01365 0 1 2 3 4 5 6 7 0 10 20 30 40 50 60 70

pH & Bare sand coverage

0 1 2 3 4 5 6 7 8 9 10

(1) Recently active (2) Stabilized (3) Wet soil (4) Valley soil

Vegetation types

A: Fully stabilized blowout B: Recently active blowout

Vegetation Type Characteristic Species

1 Viola curtisii, Corynephorus canescens, high number of lichen species

2 Carex arenaria, Hippophae rhamnoides, low number of lichen species

3 Holcus lanatus

4 Salix woods (Willow)

Figure 2. Correlation plot of pH and bare sand coverage data. The data shows no significant correlation. The slope of the trend line approaches zero, thus it is not reliable to extrapolate the data.

(32)

32

Appendix E Transect labeling key

BL1 AL1

BL2 AL2

BL3 AL3

BK19 BK18 BK17 BL4 BK16 BK15 BK14 BK13 AK19 AK18 AK17 AL4 AK16 AK15 AK14 AK13

BL5 AL5 BL6 AL6 BL7 AL7 BL8 AL8 BL9 AL9 BL10 AL10 BL11 AL11 BL12 AL12

Figure 1. Schematic overview of the blowouts and transect labeling key.

Figure 2. Actual overview of the blowouts and transect points derived from

GoogleEarth(2005).

Assessing the long-term influence of active blowout processes on soil quality in the Grey dunes of Eldorado, Terschelling