Dune slack succession
Succession in dynamic, wet dune slacks - southern Texel
Van Middelaar, J.C. Supervisor: Mw. Dr. A.M. Kooijman
Track: Future Planet Studies with Major: Earth Sciences June 29, 2015
2
Table of content
1.Introduction………..…..3 2. Methods……… Study area……….……….4 Laboratory work………..5 Statistical analysis………..6 3. Results……….……..………… Bulk density ……….……….………7 Vegetation composition……….……….9 Bare sand ……….………..10 Herb cover ……….………12 Moss cover……….……….14 Shrub cover……….………..16 4. Discussion……….……18 5. Conclusion……….………..19 6. Acknowledgements………..……..20 7. Literature………..…………21 8. Appendix………..…...25 AbstractThe southern tip of Texel contains several wet, dynamic dune slacks. These dune slacks belong to the nature protection areas of the EU Natura 2000 and contribute to the conservation and restoration of habitats and species. Also the Red List species Liparis Loeselii occurs in these young dune slacks. This orchid is a pioneer species and occurs for several years throughout the young dune slacks. After a few years it declines in number. This decline is largely triggered by succession of the soil and vegetation. With time, vegetation growth will trigger the accumulation of soil organic matter with a concomitant decrease in pH. However, by implementing Dynamic Coastal Management this development towards a more mature state can be counteracted. To create a better understanding of dune slack succession and the influence of floods on this development, the following soil and vegetation parameters will be studied: pH, electrical conductivity, organic matter content, C:N ratio, carbon content, nitrogen content, water level, bulk density and the vegetation composition. Thereby focussing on the soil bulk density and the vegetation composition.
Keywords
3
Introduction
Wet dune slacks are among the few natural ecosystems which are present in the Netherlands. Especially the dune slacks which are located in the Dutch Wadden Islands are hardly influenced by human. Due to their hydrological and geomorphological gradients, these nutrient-poor slacks provide a habitat for relative rare species (Grootjans, 2006). The endangered Liparis Loeselii (L.) Rich., is one of the rare species which occurs in the coastal dunes of the Wadden islands (Odé and Bolier, 2003). It is listed as endangered in the Bern Convention and in the Habitat Directive and has a high conservation priority (Oostermeijer, 2014). It is a short-live species which rapidly expand or disappear as conditions change (Wheeler et al., 1998).
Dune formation starts with the development of embryo dunes on the beach plain. Eventually, these dunes can be fully or partially cut off from sea. In this way primary dune slacks form. Dune slacks can also form due to blowouts of dunes until the (summer) groundwater table is reached. These slacks are known as secondary dune slacks. Generally, they form further away from the coast than primary dunes. Both slacks contain similar vegetation (Shahrudin, 2014; Grootjans, 2014). During succession, biotic and abiotic circumstances in slacks change, which causes a change in vegetation composition. The most efficient abiotic factors which determine the vegetation composition are the moisture content, pH and the availability of nutrients (Van der Craats, 2010).
Standard, natural succession in slacks takes between 20 to 30 years (Sival & Grootjans, 1996). Succession can be divided into four phases (Van der Maarel et al., 1985).
1. Pioneer phase: Mostly bare soil, with a thin layer of green algae and laminated microbial mats. Accumulation of OM is low (Van Gemerden, 1993; Grootjans, 1997).
2. Colonization of spermatophytes. Still a very low nutrient availability (Grootjans et al., 2002).
3. Development of a moss layer: Species like Pleurocarpic bryophytes, assisted by the cyanobacteria which can fix nitrogen. Invasion of tall grasses and shrubs. (Stal et al., 1994).
4. More mature stage: Replacement of non-competitive plant species by shrubs and trees. Fast accumulation of OM (Sival, 1996; Ranwell, 1972).
By slowing down the succession rate, the pioneer stage can be prolonged by approximately 5 years with the help of dune management (Grootjans et al., 2014). This can for example be done through sod cutting and mowing. However, there are also natural processes which can slow down the succession rate, such as sand shifting or floods (Spek, 2014). When slacks are overflown with water, acidification and accumulation of organic matter is counteracted. However, it can also become too wet in the dune slack. When oxygen becomes a limiting factor, toxic components can form, such as sulphides.
The aim of this research is to contribute to the current knowledge of parameters thriving succession in wet dune slacks. This study is conducted in cooperation with two BSc students: D. Stomph and H.S. Berghuis. Miss Stomp will elaborate on the Ah depth, carbon and nitrogen content and the C:N ratio. Miss Berghuis will elaborate on the electrical conductivity, height of the water level and the pH of the soil. In this research will be focused on the soil bulk density and the vegetation structures in dune slacks. This knowledge can be helpful for coastal management in the Dutch dunes, and could contribute to the restoration of dune dynamics. This research is a follow-up study of Van der Craats (2010, Jongejans (2014) and Hollaar (2014). The research question of this study is: How to create a congruent understanding of soil and vegetation succession in the wet, dynamic dune slacks of southwest Texel – focusing on the soil bulk density and vegetation structures: a field study.
Including the following sub questions:
How did the bulk density and the vegetation structures develop over time?
Which correlations do occur between soil bulk density/vegetation type and the other studied soil parameters? And which parameters can predict the soil bulk density/vegetation type?
What is the relationship of soil bulk density and vegetation structure with the occurrence of the fen orchid L. Loeselii?
4 Methods
Study area
The study area includes beaches, embryonic dunes, drift dikes, primary and secondary dune slacks and fossilized dune beach ridges and is situated at the southern tip of Texel. It is positioned in the area and surroundings of the dynamic beach plain the Hors and in the dune slacks of the Horsmeertjes. The beach plains are expanding, due to accreting shoals (Westhoff and Van Oosten, 1991).
Sandbanks form by influence of the tide regime, when the inflow of water from the North Sea during high tide neutralizes the outflow from the Wadden Sea during low tide. Due to this neutralization of the currents sand can settle to the bottom. The sandbank the Horst formed and attached to Texel in 1749. Also the sandbank The Onrust formed and attached in the beginning of the 20th century. Lastly, the sandbank the Razende Bol has formed. This bank is still moving northeast (figure 1).
During the second half of the 20th century two drift dikes and several sand dunes have been constructed which isolated a part of the beach from influence of the sea. Due to this isolation a sequence of wet dune slacks have raised and The Kreeftepolder, The Horspolder and The Horsmeertjes have been formed. Nowadays, instead of stabilizing this area with the use of drift dikes, Dynamic Coastal Management is implemented (Ecomare, 2015). Regularly this area is flooded with sea water. This happens during high tides in combination with strong winds (Spek, 2014).
The sampling locations of this research are spread over 11 dune slacks, along an northeast age gradient from south west to north east. In this series, H1 is the oldest dune slack and H13 the youngest (Figure 1; Appendix, table 1). The occurrence of the L. Loeselii is related to the age of the dune slack and the succession stage (Van der Craats, 2010). Expected is that the L. Loeselii will appear in the youngest dune slacks, the species is considered extinct in the oldest slack (H1).
Figure 1: Left panel: The formation of sand bank through the years, obtained from Google Earth (Ecomare, 2015). Right panel: Map in which the spatial distribution of the sampling plots in the dune slacks is showed.
5
The fieldwork was conducted in May, 2015. To enable comparison with previous studies, the performed field and chemical analysis were identical to the methods of Van der Craats (2010), Hollaar and Jongejans (2014). The coordinates of the sampling plots of an earlier research were used (H1 – H9)(Hollaar, 2014)(Jongejans, 2014). Two of the researched slacks are measured at two different locations: a relative low and high site (H4 and H6). Four sampling locations were added to the research. Two of them were recently formed dune slacks, in which the L. Loeselii is expected to grow (H10 and H11) and two of them were recently flooded, and are useful to study the effects of Dynamic Coastal Management (H12 and H13).
The dune slacks ages of H1- H9 were, based on air photos, determined by Van der Craats (2010). When the first vegetation was visible on the photo, from then on the slack ages. The age of H10 and H11 were, also based on air photos, determined during this research in collaboration with Miss Stomph and Miss Berghuis. The slack ages of H12 and H13 were firstly set on 16 years. However, there can also be said that they are 3 years old, as these slacks were flooded in 2012 (Spek, 2014) (Appendix, Table 1).
Field measurements Vegetation
To describe the present vegetation a sampling quadrat with a size of 1 by 1 meter was used in which the percentage of each vegetation type was determined: Open water, bare sand, mosses, herbs and shrubs. Thereafter, the vegetation height was measured.
Soil profile
By using an auger, a bore hole was made in the middle of the quadrat. Firstly, if present, the following soil horizons were described: O, Ah, C1, C2, C3. Then, based on the Munsell scale, the colour of each horizon was determined. Thereafter the thickness of each horizon was noted as well as the root depth and the total depth of the soil profile (Appendix, table 2).
Water table
Fifteen minutes after drilling the borehole, the water table was measured (it was assumed that the water table had stabilized in a quarter of an hour). This was measured as the distance between the surface and the water level. To determine the relative height of the sampling plot, the water level was also measured at the lowest site of the dune slack. To correct for daily fluctuations in the water table the water table was also measured at a reference point.
Soil sample
For each sampling plot, three soil samples were taken for chemical analysis, which comes down to a total of 180 samples (15 sampling sites * 4 plots * 3 pF rings). These samples were taken out using a pF ring of 100 cm3. The collected samples were stored in closed, plastic bags (two in the same back, one separately).
Laboratory work
Dry weight and bulk density
The soil samples were weighted (inside the plastic bags), thereby using a 2 decimal precision (correct for the weight of the plastic bag). Thereafter, the bags which contained two soil samples were, in an open bag, dried for 48 hours in the oven at 70°C. The soil samples which contained one soil sample were, also in an open bag, dried for 48 hours at 30 °C. The extremely wet samples (H6 and H3) were dried for 72 hours. Then the samples were weighted again and the bulk density could be calculated (Appendix, equation 2). Then the dried samples were crushed by hand and sieved in order to filter the material with a diameter bigger than 2 mm out. Then approximately 50-60 grams of the sieved samples was collected in numbered, cardboard boxes.
C:N Ratio
Approximately 10 gram of the sieved soil sample was grinded at 400 rpm for five minutes. Thereafter, 20-50 milligram of grinded material was put in a tin cup which was folded. For calibration, smaller folded cups with 7 milligram of sulfanilic acid were placed after, in between and for the samples. Their weight was entered as input for the CNS element analyser. Thereafter, each cup was placed individually in the combustion chamber of the analyser and burned at 1150 °C. Out of these analysis came the total percentage of carbon, nitrogen and the C:N ratio. These numbers were converted into g/m2 (Appendix, Equation 4). All samples were analysed in duplicates.
pH(H2O), pH (KCl) and EC
Firstly, 10 grams of sieved material was put in a 50 ml polyethylene bottle and 25 ml distilled water was added. Secondly, 10 gram of sieved material was put in a 50 ml polyethylene bottle and 25 ml of 1M KCl was added. For two hours, these extractions were shaken in an orbital shake machine and allowed to settle overnight. The next morning, the samples were shaken again for half an hour. Then the electronic conductivity, pH(H2O) and the pH(KCl) was measured. This was done at a temperature of 20.5°C for which is corrected during the pH measurements. These measurements were done in duplicates.
6
The statistical analysis were performed in Matlab®
Data was tested on normal distribution and equal variance using respectively the Kolmogorov Smirnov test and the Levene’s test. If the data met these assumptions a parametric test could be used. Otherwise the non-parametric test had to be performed.
Analysis methods
Development of the parameter along an age gradient
Graphs were made in which the obtained data were scattered as well as the average of each slack. Moreover, a Multiple Linear Regression was performed to test if there was a trend over time. If this trend was significant, then this regression line was showed in the graph. To test if there was a significant difference between the dune slacks a Kruskal-Wallis test was performed. Hereafter a Multiple Comparison test was used to show which years significantly differed.
Development over time: 2010, 2014, 2015
Boxplots were used to visualize the obtained data. Thereafter a non-parametric Friedman test was used to show which year differed significantly from each other. Then a Multiple Comparison test was used to check which years differed.
Correlations between parameters and with the occurrence of the L. Loeselii
The correlation between the parameters (bulk density, vegetation types: shrubs , herbs, mosses, bare sand) and the other measured parameters (C:N ratio, C content, N content, Ah depth, pH, EC, Water level) were analysed with a non-linear Spearman’s Correlation Coefficient. This correlation was performed for slacks of the same age: 3, 12, 16 and 21 years old. In order to exclude the factor time. This correlation test was also performed to test if there was a significant correlation between the bulk density/type of vegetation and the occurrence of the L. Loeselii. This correlation was used because outliers do not have a big influence of this test, because with this correlation test the data is converted to ranks. Moreover, the number of samples is relatively small and overall not normally distributed. For this reason, Spearman’s correlation is prefered over the Pearson’s correlation coefficient. To test if there was a difference in occurrence in high or low slacks the Wilcoxon Rank-sum test was performed to test the data on equal means.
Response and predictor variables
To check which parameters could predict the bulk density/vegetation type a Multiple Regression test was performed. To check if the parameter could be a reliable predictor the confidence bounds (c.b.) were checked. If the c.b. contained a zero, the parameter was considered as not reliable.
Influence of the flooding in 2012
To test if there was a significant difference between the 3 year old dune slack (H11), the 16 year old slacks (H5, H6, H7) and the overflown dune slacks (H12, H13) a one-way Anova test or a Kruskal-Wallis test was used (dependent on the distrubution and variance of the data). Thereafter a Multiple Comparison test was used which could show which slacks significantly differed from each other.
7
Results: Bulk density
Development of the bulk density over dune slack age
As shown in the graph (figure 2), the bulk density decreases significantly along the age gradient of the dune slacks: 1953 till 2012.
Figure 2. Linear regression plot of the bulk density along an age gradient of dune slacks formatted between 1953 and 2012, measured in 2015. Outlined in red are the 3 year old dune slacks of which H12 and H13 were flooded in 2012.
Development over time: 2010, 2014, 2015
The bulk density of the upper 5 cm3 of the soil decreases significantly over time. This development goes relatively fast, as there was already a significant difference measured between the data of 2010 and the data which was obtained in 2015. The measured data is shown in the boxplot below (Figure 3).
8
Correlation coefficients & Predictor-Response
There was a significant positive correlation present between the pH, Ah depth and the percentage of herb cover. And there was a negative correlation between bulk density and the percentage of moss, EC, carbon and nitrogen content and C:N ratio. It is remarkable that none of these correlations occurred in all the sampled dune slacks (Table 1).
The best parameters for predicting the soil bulk density were: carbon and nitrogen content, C:N ratio and the factor age. The nature of the relations between the bulk density and these parameters was negative.
Table 1. Spearman correlation coefficient between all soil parameters and the bulk density, measured in 2015.
pH EC Water C N C:N ratio Ah Mosses Sand Shrubs Herbs
3 0.6257 X X X X X 0.6924 -0.638 X X X 12 0.7619 X X X X X X X X X X 16 X -0.56 51 X -0.66 08 -0.62 40 -0.75 06 X X X X 0.4982 21 X -0.92 47 X X X X X X x X X Occurrence of L. Loeselii
There was no correlation between the bulk density and the L. Loeselii in 2010. But there was a significant positive correlation in 2014 between the occurrence of the L. Loeselii and the bulk density (Table 2).
Table 2. Spearman correlation coefficient between the bulk density and the occurrence of the L. Loeselii, measured in 2015.
Table X: Giving the significance of correlation and the correlation coefficient for a relation between Liparis Loesellii occurrence and the bulk density Bulk density (g/cm3) Liparis 2010 occurence r = -0.2047 p = 0.1825 Liparis 2014 occurence r = 0.4007 p = 0.0070 Influence of flooding in 2012
The mean bulk density of H12 and H13 is respectively1.4225 and 1.4450, which is relatively high compared to the other slacks. However, the soil mass of H13 does only differ significantly from the 16 year old slacks H6L and H7. There are no significant differences in bulk density between H12 and the other slacks.
9
Results: Vegetation
Figure 4 shows an overview of the vegetation composition. What is notable here is that as the dune slack are aging, the percentage of open water increases in contrast to the percentage of bare sand, which decreases over time. The percentages of mosses, herbs and shrubs vary over time and show, at first sight, no trend.
10
Results: Bare sand
Development of the percentage of bare sand over dune slack age
There is a decreasing trend in percentage of bare sand along the age gradient of the dune slacks (1953-2012)(Figure 5).
Figure 5. Linear regression plot of the percentage of bare sand along an age gradient of dune slacks formatted between 1953 and 2012, measured in 2015. Outlined in red are the 3 year old dune slacks of which H12 and H13 were flooded in 2012.
Development over time: 2010, 2014, 2015
There is a significant difference between the vegetation cover in 2010, 2014 and 2015. The measured data is shown in the boxplot below (Figure 6).
11
Correlation coefficients & Predictor-Response
There was a significant positive correlation present between bare sand and the electrical conductivity. There was a negative correlation present between bare sand and herb cover. It is remarkable that none of these correlations occurred in all the sampled dune slacks (Table 3).
The best parameters for predicting the percentage of bare sand were: Ah depth, C:N ratio and the factor age. The nature of the relations between the bulk density and these parameters were all negative.
Table 3. Spearman correlation coefficient between all soil parameters and the percentage of bare sand, measured in 2015.
pH EC Water C N C:N ratio Ah Mosses Sand Shrubs Herbs
3 x 0.5812 X X X X x x X X X
12 x X X X X X X X X X -0.8805
16 X x X x x x X X X X x
21 X x X X X X X X x X X
Occurrence of L. Loeselii
There was a significant correlation between bare sand and the occurrence of the L. Loeselii in 2014 (Table 4).
Table 4.. Spearman correlation coefficient between the percentage of bare sand and the occurrence of the L. Loeselii, measured in 2015.
Table X: Giving the significance of correlation and the correlation coefficient for a relation between Liparis Loesellii occurrence and the percentage of bare sand
Bare sand (%) Liparis 2010 occurence r = 0.0534
p = 0.7308 Liparis 2014 occurence r = 0.7748
p = 6.7235e-10
Influence of the flooding
The mean percentage of bare sand of H12 and H13 is respectively 71.2500 and 32.5000, which is relatively high compared to the other slacks. However, the soil mass of H13 does only differ significantly from the 16 year old slacks H6L and H6H. There were no significant differences in bulk density between H12 and the other slacks.
12
Results: Herbs
Development of the percentage of herb cover over dune slack age
There is no significant trend over time in the percentage of herb cover (1953 – 2012)(figure 7).
Figure 7. Linear regression plot of the percentage of herb cover along an age gradient of dune slacks formatted between 1953 and 2012, measured in 2015.
Development over time: 2010, 2014, 2015
There was a significant difference between the measured percentage of herb cover in 2010, 2014 and 2015. The measured data is shown in the boxplot below (Figure 8).
13
Correlation coefficients & Predictor-Response
There was a significant positive correlation present between the percentage of herb cover and pH, EC, Ah depth and the bulk density. There were negative correlations between the herb cover and the EC, moss cover, percentage of bare sand and shrub cover. It is remarkable that none of these correlations occurred in all the sampled dune slacks (Table 5).
The parameter which could predict the percentage of herb cover was the water level. The nature of the relations between the bulk density and the water level was positive.
Table 5. Spearman correlation coefficient between all soil parameters and the percentage of herb cover, measured in 2015.
pH EC Water C N C:N ratio Ah Bulk Mos Sand Bare Shrubs
3 0.6964 0.77 81 X X X X x x -0.85 47 X X X 12 x X X X X X X x X X -0.88 05 -0.74 54 16 X -0.70 61 X x x x 0.5388 0.4982 X X X x 21 X -0.69 47 X X X X X -0.96 30 x X X Occurrence of L. Loeselii
In 2010 and in 2014 there were no significant correlations found between the percentage of herb cover and the occurrence of the L. Loeselii (Table 6).
Figure 6. Spearman correlation coefficient between the herb cover (%) and the occurrence of the L. Loeselii, measured in 2015.
Table X: Giving the significance of correlation and the correlation coefficient for a relation between Liparis Loesellii occurrence and the percentage of herb cover
Herb cover (%) Liparis 2010 occurence r = -0.1310 p = 0.3968 Liparis 2014 occurence r = -0.1711 p = 0.2668 Influence of flood
The mean percentage of herb cover of H12 and H13 is respectively 15 and 7.5, which is relatively low compared to the other slacks. However, the moss cover of H13 does only differ significantly from the 16 year old slack H6H. There are no significant differences in bulk density between H12 and the other slacks.
14
Results: Mosses
There is a significant, increasing trend over time in the percentage of moss cover along the age gradient of the dune slacks (1953 – 2012)(figure 9).
Figure 9. Linear regression plot of the percentage of moss cover along an age gradient of dune slacks formatted between 1953 and 2012, measured in 2015. The red area marks the 3 year old dune slacks in which Bryum Spec. probably grow. Outlined in red are the 3 year old dune slacks of which H12 and H13 were flooded in 2012.
Development over time: 2010, 2014, 2015
There was a significant difference between the measured percentage of moss cover in 2010, 2014 and 2015. The measured data is shown in the boxplot below (Figure 10).
15
Correlation coefficients & Predictor-Response
There were significant positive correlations present between the percentage of moss cover and water level, EC, C:N ratio and the shrub cover. There were negative correlations between the pH, EC, bulk density and the herb cover. It is remarkable that none of these correlations occurred in all the sampled dune slacks. And that EC has a positive correlations with moss cover in the 3 year old slack and a positive correlation with moss cover in the 21 year old slack (Table 7).
The parameters which could predict the percentage of moss cover were: nitrogen and carbon content, C:N ratio, water level and the factor age. The nature of these relations were all positive.
Table 7. Spearman correlation coefficient between all soil parameters and the percentage of moss cover, measured in 2015.
pH EC Water C N C:N ratio Ah Bulk Herbs Sand Bare Shrubs
3 -0.85 56 -0.87 74 0.6040 X X X x -0.6378 -0.8547 X X 0.6003
12 x X 0.7544 X X X X x X X x x
16 X X x x 0.5516 x x X X X x
21 X 0.7452 X X X X X x -0.96 30 x X X
Occurrence of L. Loeselii
In 2010 and in 2014 there were no significant correlations found between the percentage of moss cover and the occurrence of the L. Loeselii (Table 8).
Table 8. Spearman correlation coefficient between the moss cover (%) and the occurrence of the L. Loeselii, measured in 2015.
Table X: Giving the significance of correlation and the correlation coefficient for a relation between Liparis Loesellii occurrence and the percentage of herb cover
Herb cover (%) Liparis 2010 occurence r = 0.2018 p = 0.1890 Liparis 2014 occurence r = -0.2718 p = 0.0743 Influence of flood
The mean percentage of herb cover of H12 and H13 is respectively 5 and 47.5. Which shows that the moss cover in H12 is relatively low in comparison to the other slacks. The soil mass of H13 does only differ significantly from the 16 year old slack H6H. There are no significant differences in moss cover between H12 and the other slacks.
16
Results: Shrubs
Development of the percentage of shrub cover over dune slack age
There is a significant, increasing trend over time in the percentage of shrub cover along the age gradient of the dune slacks (1953 – 2012)(figure 11).
Figure 11. Linear
regression plot of the percentage of shrub cover along an age gradient of dune slacks formatted between 1953 and 2012, measured in 2015. Outlined in red are the 3 year old dune slacks of which H12 and H13 were flooded in 2012.
Development over time: 2010, 2014, 2015
There was a significant difference between the measured percentage of shrub cover in 2010, 2014 and 2015. The measured data is shown in the boxplot below (Figure 12).
17
Correlation coefficients & Predictor-Response
There was a significant positive correlation present between the percentage of shrub cover and moss cover. There were negative correlations between the pH, EC, Ah depth and the herb cover. It is remarkable that none of these correlations occurred in all the sampled dune slacks (Table 9).
The parameter which could predict the percentage of shrub cover was the factor age. The nature of this relation was positive.
Table 9. Spearman correlation coefficient between all soil parameters and the percentage of shrub cover, measured in 2015.
pH EC Water C N C:N ratio Ah Bulk Herbs Sand Bare Mosses
3 -0.64 08 -0.62 21 x X X X -0.6310 x x X X 0.6003
12 -0.75 16 X x X X X X x -0.74 54 X x x
16 X X x x x x x X X X x
21 X x X X X X X x x x X X
Occurrence of L. Loeselii
In 2010 and in 2014 there were no significant correlations found between the percentage of shrub cover and the occurrence of the L. Loeselii (Table 10).
Table 10. Spearman correlation coefficient between the shrub cover (%) and the occurrence of the L. Loeselii, measured in 2015.
Table X: Giving the significance of correlation and the correlation coefficient for a relation between Liparis Loesellii occurrence and the percentage of herb cover
Herb cover (%) Liparis 2010 occurence r = 0.2448 p = 0.1093 Liparis 2014 occurence r = -0.0795 p = 0.6078 Influence of flood
The mean percentage of shrub cover of H12 and H13 is respectively 8.75 and 12.5. The percentages of shrub cover in H12 and H13 do not significantly differ from any other measured slack.
18
Table11.
Results of the analysis which were performed for each parameter. The p-value was set on a 0.05 significance level. In the Appendix, figure 1 is shown which ages/slacks significantly differed.
Bulk density Bare sand Herb cover Moss cover Shrub cover Difference along
the age gradient
3.1431e-07 7.9044e-09 5.6062e-06 9.1174e-05 0.0053
Difference between 2010, 204, 2014
3.3309e-06 1.7117e-05 0.0132 0.0011 0.0123
Difference between 3 and 6 year old slacks
5.0636e-09 8.6007e-04 9.9266e-04 0.0017 0.1153
*This table shows the p-values of comparison tests.
Table12.
Results of the analysis which were performed for each parameter. The p-value was set on a 0.05 significance level.
4L-4H 6L-6H Water 0.0286 0.0286 pH(h2o) 0.0286 0.9143 pH(KCl) 0.1143 0.6857 Bare Nan 0.3910 Shrubs 0.3492-> 0.2286 Herbs 0.6000 0.0857 Mosses 0.2857 0.4857 CN 0.3429 0.2000 C 0.0571 0.1664 N 0.1198 0.1808 AH 0.1143 0.4000 Bulk density 0.2286 0.0286 Liparis Loeselii (2010) 0.0286 0.0571
*This table shows the results of the analysis which were performed to test if there was a difference in mean when comparing the data from the high with the low dune slack.
19
Discussion
Development of the bulk density and the vegetation structures over dune slack age
Over time, the mass of the soil (g/cm3) decreases. This occurs due to an increase in organic matter content, which is a major modifier of bulk density (Rawls, 1983). As expected, there was a significant, negative trend measured over time for the soil bulk density.
During succession, biotic and abiotic circumstances change. Which causes a change in vegetation composition over time (Van der Maarel et al., 1985). This succession process was also shown in the studied dune slacks. Over time the amount of bare sand decreases, together with an increase in the moss and shrub cover. No trend was found for herb cover.
The trend which was found for the percentage of bare sand and shrub cover was expected. However, there were some unexpected results in the dataset of the moss cover: In the 3 year old slack there was a very high peak noticeable in moss cover. Which thereafter dropped suddenly and then started to increase again over the years. Expected is that in these relative young dune slacks the pioneer moss specie Bryum Spec grew. In contrast to the older dune slacks, in which probably a higher percentage of the species Calliergonella Cuspidata grew (Kooijman, unpublished). To gain more insight in these vegetation development it is necessary to note also the different vegetation species. Probably then could also be explained why there was no significant trend over time shown in the herb cover. However, this was not possible during this fieldwork, because the fieldwork was performed too early in the season. Thence the majority of the present vegetation was not yet recognizable. Therefore it can be preferable to perform this fieldwork later in the season (July-August), in order to gain more insight in the vegetation development.
The data from 2010, 2014 and 2015 were compared. It turned out that there was a significant difference between the parameters in 2010, 2014 and 2015. But the obtained data showed a lot of fluctuations instead of a clear trend over time (see boxplots).Therefore it can be believed that this outcome is not reliable as there is a spatial variance of the sampling sites. For further research it would be better to mark the sampled sites.
To gain insight in the vegetation characteristics of the dune slacks the vegetation height per relevée was also noted. This was done because this parameterinfluences the light availability and describe the openness of the vegetation. However, regularly the vegetation (in particular shrubs) in the measured relevée was cut. Therefore, the obtained data of the vegetation height is left out of the analyses. For further research it can be interesting to take also this parameter into account. But this can only be done in areas where there is no pruning of vegetation.
Furthermore, the sampling plots were not equally distributed through the area. Therefore there was relative little data available of old slacks. This causes difficulties in studying the development of the older dune slacks as well as the development of slacks over time. For this reason it is recommended to add a few sampling sites which are older than 25 years to create a better understanding of dune slack succession. Is there a correlation between bulk density and the vegetation composition with the other soil parameters: pH, EC, Water level, OM, Ah depth, C content, N content, C:N ratio? And can they be predicted by one of these parameters?
When comparing sampling areas of dune slacks with the same age there were several correlations found between the soil parameters. This correlation test was performed for every group of slacks with the age. But the correlations which were found differed per age group. Probably this can be explained by the relative large variation in the data. To reduce these fluctuations in the data, it can be preferable to increase the number of sampling sites per slack.
The multiple comparison test showed that the bulk density and most of the vegetation parameters could be predicted by factors which determine the organic matter content: carbon and nitrogen content, C:N ratio and the Ah depth. This was expected because it was stated by Berendse et all (1997), that the supply of nitrogen and other plant nutrients are one of the most important factors which determines the dynamic of species composition. As the amount of organic matter increases, due to the production of litter and dead roots, the soil moisture increases and the soil pH declines. These changes in soil features have an impact upon the establishment of plant species. The parameter water was also a significant predictor for the percentage of moss and herb cover. This is expected as water is a major modifier of vegetation composition. Therefore it was unexpected that water was no predictor for the other two vegetation parameters. Probably this is because mosses and herbs have relative short roots and are therefore very vulnerable for fluctuations in the water table. The parameter age is the best predictor for the percentage of shrub cover. In the dune area the increased atmospheric deposition and a lowering of the groundwater table increased the percentage of shrub cover (Salix Repens). This increase in shrub cover accelerates the increase in organic matter accumulation and thereby causes a faster decline of some character species (Bakker et al., 1999).
What is the relationship of soil bulk density and vegetation structure with the occurrence of the fen orchid Liparis Loeselii?
The populations of the L. Loeselii do rapidly expand or disappear as conditions change. The characteristic habitat of this orchid is formed by open vegetation on nutrient-poor, base-rich or calcareous, wet soils (Kreutz and Dekker, 2000). Therefore a relation between the bulk density/percentage of bare sand and the orchid was expected.
20
It was expected that there was a significant difference in water level, OM content, pH and the number of L. Loeselii between the relative low and high locations (H4 and H6). This was expected because these nutrient poor and base rich water can enhance preferable living conditions for pioneer species. However, these differences were not significant for the pH(H2O), OM and the EC. In contrast to the expectations the orchids preferred to grow in the high located sites. Probably this is because they cannot stand longtime inundation. Their root system would die due to the anaerobic conditions. Only plants with well-developed air tissues (Aerenchym) can survive these wet episodes, like the Schoenus nigricans (Slings et al, 2006).
What is the influence of a flood?
To protect the land behind the fore dunes, Ammophila arenaria has been planted by man since 1850. As a result of this stabilization, the dynamics of the Dutch dune reduced (Arens and Wiersma, 1994). In order to provide a suitable habitat for pioneer species it is recommended torestore the dune dynamics. Thereby creating a dynamic mosaic of vegetated and bare patches (Arens et al., 2009). It is stated by Grootjans (2006), that this can be done with Dynamic Coastal Management. Thereby using wind and water to distribute the sand which is washed ashore.
It is stated by Van der Spek (2014) that part of the sampling area is regularly overflown with sea water. Our supervisor stated that the locations H12 and H13 were overflown during 2012 (Kooijman, unpublished). During this flood a sand layer was deposited. In H12 this layer had a thickness of approximately 2 cm. In H13 this sand layer had a thickness of approximately 4 cm.Therefore it was expected that the soil conditions in this area would be influenced and rejuvenation of the soil would have occurred. By comparing the data of the bulk density it was notable that the bulk density was more comparable to the bulk density in the 3 year old slacks than the density in the 16 year old slack. However, this was only significant for slack H13. This is probably because the flood had a bigger sand layer deposited in H13, than in H12. When comparing the vegetation types there were no remarkable differences seen. However, in the field was noticed that the shrubs which grew in the overflown plots were much younger than in the 16 year old plots. Thus there is a difference, although this is not seen in our vegetation data.
That managers change their strategy from static to dynamic management is, especially in the future, of great importance. For it is necessary that the island can raise in altitude as the sea level is rising (Nichols, 2010). This rise in altitude can happen when wash-overs reach the area and thereby deposit sand.
The influence of the flood in 2012 on pioneer species such as the L. Loeselii is not known yet. However, this data will be available soon (Kooijman, unpublished). It can be interesting to keep sampling these overflown plots the coming years in order to gain insight in the influence of a flood on the soil and vegetation properties. In this way the impact the impact of different management regimes can be studied.
Microbial mats
The effects of microbial mats are not clear yet. On the one hand states Kooijman (unpublished) that succession is reinforced by algal mats, since they stabilize the slacks and provide the first material for organic matter and nutrients. On the other hand is stated by Grootjans (2002) that development towards a more mature state is counteracted by algae and cyanobacteria. According to him are microbial mats important stabilization mechanisms which can slow down erosion in dune slacks. By inhibiting the growth of species which occur in later succession states, like the Calamagrostis epigejos, they extend the life span of early pioneer species. Moreover, algae consume CO2 which contributes to the precipitation of CaCO3 (increasing the pH). In our sampled plots microbial mats occurred only in H11 and H13. However, there were no significant differences in the mass of the bulk density and the vegetation composition of slack H12 with H11 and H13. No conclusions can be drawn from these results. As these researched group is too small. Further research needs to be carried out, in order to gain a complete understanding of the effects of microbial mats on dune slack succession.
Conclusion
In line with previous studies bulk density decreases over time. The bulk density can best be predicted by parameters which are representatives for the soil organic matter: C:N ratio, nitrogen and carbon content. Moreover, it can be predicted by the factor age. Along the age gradient, the percentage of bare sand and moss cover increased, together with a decrease in percentage of shrub cover. There was no significant trend for the percentage of herbs. The vegetation types could be predicted by several parameters which make up of the soil organic matter, the moisture content and also by the factor age.
In 2014, a positive correlation between soil bulk density and the percentage of bare sand with the occurrence of the L. Loeselii was measured. This showed that a decrease in soil bulk density and in the percentage of bare sand can form a threat to the occurrence of pioneer species. Preferable soil conditions as a relatively high pH and a relative low organic matter content, as a consequence of a flood, were also seen in the research of Berghuis and Stomph (2015). Therefore, an increase in dynamic processes seems to be preferable for pioneer species.
21
Acknowledgements
I would like to thank Mw. Dr. A.M. Kooijman for her supervision during the whole project, Dhr. L. Hoitinga and Ms. J.C. Schoorl for their guidance during the laboratory work, Dhr. Dr. W.M. de Boer for his guidance concerning the GPS device and ArcGIS and Dhr. Dr. Ir. E.E. van Loon for his guidance and recommendations concerning prudent statistical analyses.
22
Literature
Van den Ancker, J.A.M., Jungerius, P.D. & Mur, L.R. (1985): The role of algae in the stabilisation of coastal dune blow-outs. Earth Surface Processes and Landforms 10, 189-192.
Arens, S.M., and Wiersma, J. (1994). The Dutch Foredunes - Inventory and Classification: Journal of Coastal Research, v. 10, p. 189-202. Arens, S.M., Van den Burg, A.B., Esselink, P., Grootjans, A.P., Jungerius, P.D., Kooijman, A.M., De Leeuw, C., Löffler, M., Nijssen, M., Oost, A.P., Van Oosten,H.H., Stuyfzand, P.J., Van Turnhout, C.A.M., Vogels, J.J., and Wolters, M., 2009,Preadvies Duin- en Kustlandschap, Directie kennis, Ministerie van Landbouw,
Natuur en Voedselkwaliteit. Arens, S.M., and Wiersma, J., (1994) The Dutch Foredunes - Inventory and Classification: Journal of Coastal Research, v. 10, p. 189-202.
Berendse, F., Lammerts, E. J., & Olff, H. (1998). Soil organic matter accumulation and its implications for nitrogen mineralization and plant species composition during succession in coastal dune slacks. Plant Ecology, 137(1), 71-78.
Berghuis, H.S. (2014). Soil development and interactions in dune slacks of the dynamic coastal area De Hors, Texel: A field study concerning the acidity, electrical conductivity and water table. Bsc Thesis. Free University of Amsterdam
Bodemacademie (2014). Parameters. Received at June 24, 2015. From source: http://www.bodemacademie.nl/index.php?i=318
Craats, A. (2012). Window of opportunity of Liparis loeselii in relation to dune slack development on Texel. M.Sc. thesis, IBED, UvA, pp. 1-116.
Ecomare (date unknown). Natuur op Texel. De Hors. Reiceived at April 18, 2015. From source: http://www.ecomare.nl/ecomare-encyclopedie/gebieden/waddengebied/nederlands-waddengebied/texel/natuur-op-texel/hors/
Grootjans, A.P. et al. (1991) Succession and fluctuation in a wet dune slack in relation to hydrological changes, J. Veg Sci. 2, 545-554 Grootjans, A. P., Geelen, H. W. T., Jansen, A. J. M., & Lammerts, E. J. (2002). Restoration of coastal dune slacks in the Netherlands. In Ecological Restoration of Aquatic and Semi-Aquatic Ecosystems in the Netherlands (NW Europe) (pp. 181-203). Springer Netherlands. Grootjans, A. P., F. P. Sival, and P. J. Stuyfzand. (1996). Assessingthe restoration perspectives of degraded dune slack ecosys-tems: a geohydrochemical approach. Vegetatio 126:27–38.
Grootjans, A., Stichting, E. R. A., Stuyfzand, P., Petersen, J., & Shahrudin, R. (2014). Ontwikkeling van zoet-zoutgradiënten met en zonder dynamisch kustbeheer.
Hollaar, T.P. (2014). Exploring the influence of succession in wet coastal dune slacks on pioneer species Liparis Loeselii, with focus on soil organic matter and nutrient content along an age gradient. BSc Thesis. Free University of Amsterdam.
Jongejans, L. (2014). Exploring the change in succession in dune slacks in the Hors, Southwest Texel between 2010 and 2014, focusing on the acidity and electrical conductivity of the soil and the water table level: a field study. BSc Thesis. Free University of Amsterdam. Kreutz, C.A.J., Dekker, H., 2000. De Orchidee€en van Nederland. Ecologie, Verspreiding,
Bedreiging, Beheer. Seckel and Kutz, Raalte/Landgraaf.
McCormick, M.K., Jacquemyn, H., (2014). What constrains the distribution of orchid populations? New. Phytol. 202, 392e400.
Nationaal Park Duinen van Texel (date unknown). De zuidpunt van Texel. Received at April 18, 2015. From source: http://www.npduinenvantexel.nl/Natuurgebieden/De-zuidpunt-van-Texel
23
Nicholls, R. J., & Cazenave, A. (2010). Sea-level rise and its impact on coastal zones. science, 328(5985), 1517-1520.
Nordstrom, K. F., & Arens, S. M. (1998). The role of human actions in evolution and management of foredunes in The Netherlands and New Jersey, USA. Journal of Coastal Conservation, 4(2), 169-180
Odé, B., and Bolier, A. (2003), Groenknolorchis op de kaart: Gorteria, v. 29, p. 33-37.
Oostermeijer, J. G. B., & Hartman, Y. (2014). Inferring population and metapopulation dynamics of Liparis loeselii from single-census and inventory data. Acta Oecologica, 60, 30-39.
Pillon, Y., Qamaruz-Zaman, F., Fay, M. F., Hendoux, F., & Piquot, Y. (2007). Genetic diversity and ecological differentiation in the endangered fen orchid (Liparis loeselii). Conservation Genetics, 8(1), 177-184.
Ranwell, D. S., (1972). Ecology of Salt Marshes and Sand Dunes. Chapman and Hall.
RAWLS, W. J. (1983). ESTIMATING SOIL BULK DENSITY FROM PARTICLE SIZE ANALYSIS AND ORGANIC MATTER CONTENT1. Soil Science, 135(2), 123-125.
Shahrudin, R. (2014). Do we really need management to preserve pioneer stages in wet dune slacks? (Doctoral dissertation, PhD Dissertation University of Groningen, Groningen).
Sival, F. P. (1996). Mesotrophic basiphilous communities affectedby changes in soil properties. Acta Botanica Neerlandica 45:119–131. Sival, F.P. 1996. The buffercapacity and related soil forming processes of two dune slack succession series. Acta Bot. Neerl. 45: 95-106 Stal, L. J., M. Villbrandt & B. de Winder. (1994). Nitrogen fixation in microbial mats. In Krumbein, W. E., D. M. Paterson & L. J. Stal (eds), Biostabilisation of Sediments. BIS Verlag, Oldenburg, FRG: 384–399.
Slings, R., Grootjans, A., Brouwer, E., (2006) Aptroot, A. (2006). Natuurkennis: Kalkrijke duinvalleien. Received at 16-06-2015. From source:
http://www.natuurkennis.nl/index.php?hoofdgroep=2&niveau=3&subgroep=107&subsubgroep=1022&subsubsubgroep=525 Spek, E. (2014) Storm over De Hors en Mokbaai op Texel. Received at 17-06-2015. From source:
https://staatsbosbeheertexel.wordpress.com/2014/10/22/storm-over-de-hors-en-mokbaai-op-texel/
Spek, E. (2014) Hoog zeewater door harde wind op Texel goed voor natuur en natuurgenieter. Reiced at 17-06-2015. From source: https://staatsbosbeheertexel.wordpress.com/2015/01/11/hoog-zeewater-door-harde-wind-op-texel-goed-voor-natuur-en-natuurgenieter/ Stal, L.J., van Gemerden, H. & Krumbein, W.E. 1985. Structure and development of a benthic marine microbial mat. FEMS Microbiol. Ecol. 31: 111-125.
Stomp, D. (2014) How soil development relates to succession- the young wet dune slacks, southern Texel. Bsc Thesis. Free University of Amsterdam.
Stuyfzand, P. J. (1993). Hydrochemistry and hydrology of theCoastal Dune area of the western Netherlands. Ph.D. thesis. Free University of Amsterdam.
Van der Maarel, E., R. G. A. Boot, D. Van Dorp, and J. Rijntjes. (1985). Vegetation succession on the dunes near Oostvoorne,The Netherlands: a comparison of the vegetation in 1959 and1980. Vegetatio 58:37–187.
24
Westhoff, V. (1989). Dunes and dune management along the NorthSea coasts. Pages 41–51 in F. van der Meulen, P. D. Junge-rius, and J. Visser, editors. Perspectives in coastal dune man-agement. SPB Academic Publishing, The Hague.
Westhoff, V., and Van Oosten, M.F. (1991), De plantengroei van de Waddeneilanden: Utrecht, Stichting Uitgeverij Koningklijke Nederlandse Natuurhistorische Vereniging, 417 p.
Wheeler, B.D., Lambley, P.W., Geeson, J., (1998). Liparis loeselii (L.) Rich. In eastern England: constraints on distribution and population development. Bot. J. Linn. Soc. 126, 141e158
25
Appendix
Table 1: The year in which the first vegetation was noticeable was stated as the year in which the dune slack formation has ended (Van der Craats, 2010).
* Originally these slacks are formed in 1999. However, after a flood in 2012, new sand has been deposited (+/- 5cm) and rejuvenation of soil and vegetation has occurred.
Location Code No. of
sampling plots Date of first vegetation First occurrence of L. Loeselii Dune type N-Nw of western Horsmeertje H1 4
1970 1978 Primary dune slack
Grauwe Gans valley H2 4
1986 1992 Secondary dune slack
Kreeftepolder East H3 4
1994 1995
Primary dune slack Kreeftepolder middle – low H4L 4 Kreeftepolder middle – high H4H 4 Secondary dune slacks north of Horsvallei H5 4 1999 1998
Secondary dune slack
Horsvallei – low H6L 4
Primary dune slack
Horsvallei - high H6H 4
Saltier part
Horsvallei H7 4
Western new dune
slack H8 4 2003 2008 Future slack in Horsvallei H9 4 2009 X H10 4 2003 2014 H11 4 2012 X H12* 4 H13* 4 Total 60
26 Soil horizons
O When a peat layer was present (>50% organic matter)
Ah When an organic horizon was visible (percentage of organic matter < percentage of sand) C1 The permanent oxidized substratum (no mottling: brown color)
C2 The both oxidized and reduced substratum (with mottling)
C3 The permanent reduced substratum (barely/no mottling: grey/blue color + smells of sulphide) Table 2 *Source: Van der Craats, 2011.
Equations
3). Hereby is a the consumption of HCL at the last inflection point
4). From the results of the CNS analysis, the average between the two measurements per sample for total percentage carbon and nitrogen was calculated. Percentages were then converted into mmol/kg according to formula 4, with b= concentration of the measured parameter, and c= atomic mass.
5). Concentrations measured in the water extracts were converted from micromole/L to mmol/kg with formula 5. With b= concentration of the measured parameter, d= moisture loss at 105 ºC and f= 105 ºC oven dry sample.
6).Concentrations measured in the water extracts in ml/L were converted to mmol/k with formula 6. With b= concentration of the measured parameter, c= atomic mass, d= moisture loss at 105 ºC and f= 105 ºC oven dry sample.
7). For all parameters the concentration in mmol/kg was converted into concentration in mmol/m2, which incorporated to 1 m by 1 m vegetation relevèe and the 5
cm depth of the soilsample (depth of pF-ring). This was done with formula 7. With b= concentration of the measured parameter. The concentration in mmol/m2 is
the relevèe topsoil density to a depth of 5 cm. *Source: Van der Craats, 2011.
27
28
Figure 2. Results of the Multiple comparison tests for soil percentage of bare sand
29
30
