The impact of climate change on the regrowth of disturbed heathland A manipulation experiment to identify changes in heathland species dynamics.

(1)

The impact of climate change on the regrowth of

disturbed heathland

A manipulation experiment to identify changes in heathland species dynamics.

Stef Knibbeler 10633316, UvA Supervisor: Albert Tietema Date: 03-07-2016

1

Knibbeler, S. (2016). ‘The impact of climate change on the regrowth of disturbed heathland’ University of Amsterdam.

(2)

Samenvatting:

Op de klimaattop in Parijs (2015) werd eens temeer benadrukt dat klimaat verandering niet langer kan worden genegeerd. De huidige klimaat problematiek is urgent en dient deze eeuw te worden verholpen. Steeds meer kennis over de impact van klimaat verandering draagt bij aan de adaptatie en mitigatie van onze hedendaagse maatschappij. Echter, is deze vorm van onderzoek vaak gericht op de impact van klimaat verandering op onze ecosystemen. Een belangrijk aspect wat hierbij achterwege dreigt te geraken is het herstelvermogen van deze ecosystemen. De gevolgen van klimaatverandering dreigen hierdoor ernstig onderschat te worden. In dit onderzoek wordt klimaatverandering nagebootst op gemaaid heideterrein gelegen nabij het Harde in de Veluwe. In deze studie is de impact van deze temperatuur- en neerslagmanipulatie onderzocht op vegetatie niveau. Uit dit onderzoek is naar voren gekomen dat verhoogde temperaturen de groeisnelheid van struikheide positief beïnvloeden. Droogtes, daarentegen laten een tegengesteld effect zien. Ook is er een mogelijk verband aangetoond tussen een verhoogde temperatuur en een verhoogde fotosynthese. Of deze effecten de stabiliteit van heide verbeteren of aantasten is niet met zekerheid te zeggen. Feit is dat relatief kleine klimatologische veranderingen het herstelvermogen van heide kunnen beïnvloeden. Dit

onderstreept het belang van onderzoek naar deze specifieke impact van klimaat verandering om zo in de toekomst onze ecosystemen intact te houden.

3

(4)

Abstract:

The knowledge about the impact of climate change on our ecosystems is increasing all over the world. More and more is known about the interactions and feedbacks of the earth system. But climate change not only affects our ecosystems itself, it also affects the capability of ecosystems to recover from natural and anthropogenic disturbances. The impact of climate change could be

dramatically underestimated if not accounted for these effects. The aim of this study is to identify the impact of climate change on a recovering heathland vegetation in Oldebroek, the Netherlands. To study these effects a manipulation experiment is conducted to identify the impact of these changing weather patterns. In this study the effects of the manipulation experiment are analysed by looking at vegetation dynamics in a disturbed heathland ecosystem and those will be compared with previous findings to learn more about the rate of changing vegetation dynamics. Results of this study show that manipulating temperatures and droughts are likely to affect regeneration of heathland species by plant productivity and species composition, which underlines the importance of understanding how climate change affects the recovery capacity of ecosystems.

Key words: Heathland, manipulation, climate change, vegetation dynamics, NDVI

(5)

Introduction:

According to the Intergovernmental Panel on Climate Change (IPCC, 2013) temperatures and precipitation levels in the Netherlands are expected to increase in the 21th_{century. More precisely:}

the annual mean temperature is expected to increase with 3-3.5 °C and precipitation levels are expected to increase with 5-10 % between 2071-2100. However, the predicted changes show a high regional variability. Figure 1 shows this variability in Europe. To account for this variability seven experimental sites were constructed throughout Europe to create temperature and precipitations gradients(Ransijn, 2014). According to Kröel-Dulay et al. (2015) European shrub lands seemed to be quite resistant to moderate warming and drought experiments at these experimental sites. However, disturbances in these shrub lands are likely to modulate climate change: Regeneration of traditional species after a disturbance can be affected by climate change that would otherwise hardly affect the mature vegetation. Therefore Kröel-Dulay et al (2015). concluded that the impact of climate change could be underestimated and the sensitivity of an ecosystem could be highly dependent on its ability to recover from a disturbance. To confirm these conclusions in this study the impact of climate change on the regrowth of disturbed heathland is investigated in the study area of the Netherlands. Furthermore earlier obtained data by Schaap et al. (2015) will be compared with results from this study to create a longer timeframe to detect significant vegetation shifts. Understanding these shifts can be used to identify appropriate time steps to monitor vegetation patterns.

5

(6)

History

The aim of the study is to investigate the impact of changing climate conditions on disturbed heathland. The Heathlands in the Netherlands can be considered as a typical form of a cultural landscape which means that they were formed (un)intentionally by human activity. Farmers in the late middle ages used distant areas from villages for their sheep’s to graze and the produced manure was used on agricultural fields. This type of field management caused low nutrition levels in these soils which were ideal for heathland species to grow. By maintaining this low nutrition levels for several decades heathland could develop. With the invention of artificial fertilizers the need for manure decreased and heathland could be used for agriculture. Furthermore a large fraction of heathlands in the Netherlands was replaced by forests. Nowadays most heathlands have

disappeared, the remaining heathlands are maintained by environmental agencies for recreation and for ecological diversity (Geologievannederland.nl).

Research question:

Worldwide the knowledge about the impact of climate change on our ecosystems is increasing extremely fast. However the impact of climate change often has a steady state approach which means that the effect of climate change is only studied when undisturbed. Because of this unilateral approach the impact of climate change on our ecosystems could be radically underestimated(Kroël-Dulay et al, 2015). Therefore in this study the impact of climate change is studied in combination with human disturbances. How does an ecosystem regenerate and recolonize after a disturbance under the stress of a changing climate. Heathlands have a relatively low species richness and therefore changes in these ecological structure are relatively easy to understand. This makes the heathland ecosystem a suitable area for studying these effects(Ransijn,2014). This results in the following research question:

To what extent does climate change affect the regrowth of disturbed heathland vegetation after mowing ? To answer this question the following sub questions need to be answered:

1. What is the effect of manipulating temperature and precipitation on the regrowth of heathland vegetation?

2. How does the vegetation cover change in one year?

3. What is impact of climate change on the photosynthetic capacity of heathlands? 4. How will heathlands develop in the future?

(7)

Hypothesis:

 What is the effect of manipulating temperature and precipitation on the regrowth of heathland vegetation?

Ericaceous species, like the Calluna vulgaris prevent nitrogen mineralization to maintain reduced nutrient levels in which the Ericaceous family favours (Latham, 2003). Increasing temperatures are expected to increase microbial activity in the soil which makes it more difficult for Ericaceous species to suppress nitrogen mineralization which results in favourable conditions for other species such as the Deschampsia flexuosa which favours higher nitrogen levels. Therefore it is expected that Ericaceous species are replaced by other grasses and mosses because of warming.

The effect of droughts can be predicted by distinguishing plant species into two categories: Low flexibility species and opportunistic species(Grubb, 1998). Low flexibility species have long-lived leaves, a relatively slow growth rate and a relatively low root-shoot ratio whereas opportunistic species reduce respiration rate during stress events , have relatively high growth rates and a high root shoot ratio. Calluna vulgaris is a low flexibility species and Deschampsia flexuosa can be considered as an opportunistic species (Grubb, 1998). Extreme droughts could be beneficial for Calluna vulgaris because of its relatively high stress tolerance compared to grass species (ibid.). Figure 2 shows the difference between opportunistic and low flexibility species. Stress events (such as extreme droughts) seem to have a lower impact on the photosynthetic capacity of

flexibility species. Therefore it is hypothesized that the vegetation cover will be dominated by Calluna vulgaris.

 How does the vegetation cover change in one year? According to Kroel Dulay et al(2015). Heathlands are quite resistant to experimental warming and droughts when undisturbed. So therefore it is hypothesized that vegetation dynamics are relatively stable in the perspective of one year. On the other hand it is hypothesized that the rate of change in vegetation dynamics is larger in the recently (2013) disturbed

plots because of the expected relatively high abundance of opportunistic species as discussed in the previous section. 7

Figure 2: Season pattern of opportunistic species(green) and low flexibility species (purple). The black square indicates a stress event. (Ransijn, 2014).

(8)

 What is impact of climate change on the photosynthetic capacity of heathlands?

The limiting factor for C3 plants for photosynthesis is the role of RuBisCO. If CO2 Concentrations are

too low oxygen binds to RuBisCO which leads to photorespiration causing some of the energy produced by photosynthesis to be wasted. Therefore C3 plants will benefit from higher CO2 levels. According to (Niinemets, Flexas & Peñuelas 2011) there is a stronger response to elevated CO2 levels

in terms of water use efficiency (WUE) in species with a more robust structure(hard leafs and short distance between leafs along the stem) such as the Calluna vulgaris. Therefore it is hypothesized that the photosynthetic capacity of Calluna vulgaris does not significantly differ between the drought and the control plots. As for the warming plots an increase in photosynthetic capacity is hypothesized. This is because according to NASA the limiting factor for plant productivity in Western Europe is a combination of solar radiation and temperature(Figure 3). Therefore it is expected that night-time warming increases plant productivity and the photosynthetic capacity.

 How will heathlands develop in the Future?

(Mobaied et al. 2010) attempt to predict the vegetation dynamics of heathland vegetation using a diachronic spatial approach. The potential future vegetation was modelled using Markov Chains coupled to a GIS Program. The model predicted a gradual change in favour of Molinia caerulea at the expense of calluna Vulgaris. In this study is hypothesized that increased nitrogen deposition

influenced the afforestation of the heathland areas. For the area of Oldebroek a similar trend is 8

Figure3: Limiting factors plant productivity (Image by Robert Simmon NASA Earth Observatory, based on data provided by the University of Montana NTSG)

(9)

hypothesized: It is expected that nutrient scarcity adapted species will be overgrown by less adaptive species in the warming plots because of increased microbial activity.

Methodology

Research area

To answer the research question and sub questions a field experiment will be conducted at the Dutch heathland research location near Oldebroek. At the location there are nine plots that will be

investigated. At three plots droughts are simulated, three other plots are warmed at night time and the remaining three plots represent the control states. Figure 4 visualizes the location of these plots. Each plot is approximately 20 m2_{of which 50% is undisturbed heathland, 40% was cut in 2013 and 10}

% was cut in 2009. The Drought plots react to precipitation by a sensor that closes at a specific amount of rainfall. This system reduced the annual soil moisture content with 33% and during

drought events with 82% (Beier et al., 2004). The amount of water that is not intercepted by the curtain and reaches the soil is continuingly measured. A similar system is used to increase the soil temperature. A light intensity sensor was used to close the curtains during night-time. These curtains intercept reemitted longwave radiation by the earth’s surface which led to an air temperature increase of 0.7 degrees and a soil temperature increase of 0.4 degrees during night time(ibid.). To keep the hydrological conditions unaffected in these plots the curtains were opened during a rain event at night. In each of these plots the disturbed part will be mapped. Figure 5 gives an impression of how a plot looks like.

9

Figure 4: Location of the plots. (Kopittke et al. , 2013)

(10)

Answering sub questions:

 What is the effect of manipulating temperature and precipitation on the regrowth of heathland vegetation?

This research question will be answered by mapping the vegetation of all 9 plots. For each plot the vegetation types will be determined by accurately map the vegetation cover in the 2009 and 2013 plots (figure 5). Each plot will be 2 m2_{. The criteria for bare soil is set on less than 20 percent}

vegetation cover. This resulted in a fraction of total area covered in a plot for each species. Each species contained three fractions for a specific treatment and year cut (For example: Calluna vulgaris C1 = 43% Calluna vulgaris C2 = 32% Calluna vulgaris C3 = 38%) . For each species in a specific

treatment, a one way ANOVA will be conducted to determine whether the year cut, treatment type or a combination of these two factors significantly influences the species cover.

 How does the vegetation cover change in one year?

To answer this research question the vegetation cover of each species found by Schaap (2015) will be compared with obtained data from this research. Combining both datasets result in a timeframe of 5 years which would be a sufficient indication to identify vegetation shifts. The timeframe will include the vegetation cover of all species 2,3,6 and 7 years after the plots were cut. To combine these two datasets it is vital both so not show extreme random variations. Therefore the average species cover 10

(11)

of all 9 plots will be compared using a two sided T-test to detect general vegetation shifts in the area. This will be done by comparing the cover of each species in a specific treatment with the dataset from 2015. For example: The mean cover of Calluna vulgaris in all control plots will be compared with the mean cover of Calluna in control in the 2015 dataset, to detect significant changes between both datasets. If the datasets are comparable the timeframe of 5 years will be used to calculate the average change per year of each species in a treatment.

To answer this research question a NDVI (Normalized Difference Vegetation Index) measurement device will be used to measure the abundance of vegetation. This device measures the ratio of Infrared light and visible light (large fraction of the visible light is used by plants for photosynthesis). The NDVI ratio is calculated as followed: The device measures the difference between Near Infrared Radiation(NIR) and Visible light(VIS) and this is divided by the total incoming radiation. This results in a NDVI ratio that ranges from -1 to 1 and values above approximately 0.2 indicate that vegetation occurs(NASA,2000). These values will be compared to detect differences in photosynthetic capacity between the different plots. Furthermore the absorbance of the wavelengths 507 and 590 nm will be compared to identify whether changes in precipitation and temperature have different effect on these wavelengths. The device measures the NDVI per 0.5 m2_{. Therefore each plot had 4}

measurements. (Table 6)

Results:

11

(12)

Table 1a/1b in the Appendix shows the mean cover and standard deviation of all species for the categories treatment and year cut. In this section the key results of this table will be highlighted. First of all the fraction of bare area shows the largest decrease in the control plots from 64.1%(s = 19.2%) to 42.7%(s = 23.9%) between the 2013 and 2009 plots. Cladonia covered 0.2%(s= 0.4 %) of the drought plots cut in 2009 and covered 29.2% (s= 21%) of the drought plots cut in 2013. Dicranum scoparium decreased in all treatments but showed the largest decrease in the temperature plots: 31% (s = 12.9%) to 4.9% (s = 1.6%). Besides, it is notable that Cladonia and Dicranium scoparium are practically not abundant in the 2013 plots. Hypnum jutlandicum also decreased in all treatments and showed the largest decrease in the drought plots: 18.1% (s = 14.6%) to 1.8% (s = 1.6%).the fractions of Poaceae did not show significant differences between type of treatment and year cut. Polytrichum increased in the control plots from 0.4% (s = 0.4% ) to 4.3% (s = 0.6%) and decreased in the drought plots from 2.1% (s = 1.8%) to 0.1% (s = 0.1%). Rumex acetosella is practically absent in the control and temperature plots but in the drought plots do have a cover of 4.2% (s = 3.7%) in 2009 and 5.6% (s = 6.5%) in 2013.

To understand which factors determine the species composition, table 2 (Appendix) shows the resulting p-values from a Two-Way ANOVA analysis. The p-values show to which extent a fraction of a certain species is determined by type of treatment, year cut or by an interaction of these two

categories with a significance level of p<0.05. Considering the p-values for bare, it can be concluded that the differences in cover are influenced by

type of treatment and year cut. There seems to be no interaction between these categories which indicates that a single combination of a specific year and treatment does not

significantly differ from the other possible combinations. This however does seem to be the case on the species composition of Calluna vulgaris. Figure 6 shows there are significant differences in Calluna cover between the control and temperature plots regardless of the year cut. Looking at the p-values in table 2 these differences can be

12

Figure 6. Cover of Calluna vulgaris in different treatments (C = Control, D = Drought and T = Temperature)

(13)

explained by type of treatment and year . Furthermore the p-value for interaction is relatively low. Although this value is above the significance level ,(p = 0.07) it could indicate that a combination of treatment and temperature determines the cover. The increase of Calluna vulgaris between 2009 and 2013 is approximately similar in the control and drought plots. The increase of Calluna vulgaris is significantly larger in the temperature plots. Results from a multiple comparison of this data showed that all combinations of year and treatment were significantly different(p < 0.05) from T2009 except for the C2009 plots (p= 0.066) which is still a relatively high significance.

The most species have a p-value below the significance level for the category year. This means that the species cover is determined over time. Only Poaceae, Rumex and Polytrichium do not change significantly over time (figure 7B).

Time series:

Furthermore the data was compared with earlier obtained data by Schaap (2015) . Most species showed no significant change which could indicate for a relatively stable system. To detect changes, the mean fraction of each species within a specific treatment was compared with the data from 2015 using a two-sided T-test. The result of this comparison can be found in the Appendix in table 3. Only

13

A

B

Figure 7: Species cover(%) per treatment(A) , year cut (B)

(14)

the mean cover of bare area in the drought plots encountered a significant change(p = 0.035). Furthermore the change of Calluna vulgaris in the drought and control plots and the change of Polytrichum showed significance levels below (p = 0.10). all other species remained relatively stable, which means both datasets are comparable and can be used to create a time series( figure 8). The data in this research represent the fractions of 7 and 3 years after the plots were cut, in 2009 and 2013 respectively. Schaap (2015) collected data 6 and 2 years after the plots were cut. Combining these two datasets resulted in a timeframe of 5 years in which the average change per year was calculated for all species (figure 9). Table 4 shows the change per year with the standard deviation. Calluna Vulgaris increased with 3.4% (s = 2.8%) per year in the control plots, with 2.0%(s = 3.4%) in the drought plots and 7.0% (s = 2.3 %) in the temperature plots. Campylopus introflexus also

increased in all treatments with 1.8% in the control(s = 1.7%) and drought(s = 2.7%) plots and 1%(s = 2.9%) in the temperature plots. Cladonia showed no significant changes in the control and

temperature plots with an increase of 0.8%(s = 1.0%) and 1.5%(s = 4.3%). However, in the drought plots Cladonia increased with 4.9%(s = 3.4%). The change of Dicranum scoparium showed high variability in the control(0.56% s = 4.1%), drought(2.3% s = 3.4%) and temperature (-0.9% s = 6.1% ) plots. The moss Hypnum jutlandicum decreased in all different treatments: -1.7% (s = 2.3%) in the control plots, -3.5%(s = 2.1% ) in the drought plots and -2.0%(s = 2.9%) in the temperature plots. Poaceae remained relatively stable in the control(0.19% s = 1.0%), drought(0.0% s = 1.3%), and temperature plots(0.25 % s = 0.9%). Polytrichum showed a similar pattern with a change of 0.7% (s = 0.4%) for control, -0.3%(s = 1.9%) for drought and 0.6% (s = 1.7%) in the temperature plots. Rumex acetosella which is not abundant in the control plots increased with 0.3%(s = 2.4%) in the drought plots and decreased with -0.7% (s = 1.4 %) in the temperature plots. The fraction of bare area decreased in all different treatments with -5.8% (s = 7.0%) for control, -7.6% (s = 4.1%) for drought and -7% (s = 6.3%) in the temperature plots. It can be noticed that a large fraction of the species have a high standard deviation indicating that the changes have a relatively large variation

(15)

Figure 9: Species Development (%) in control, drought and temperature plots. 15

(16)

NDVI

First of all, no significant differences were found comparing the effect of treatments on the NDVI-570 and NDVI-509. The effect of the different treatment types showed a similar pattern for both

wavelengths. The differences are visualized in figure 10. The values for NDVI-509 are higher but seem to react the same on night-time warming and controlled droughts. The difference between type of treatment however does seem to be significant by looking at figure 10 . A two-sided T-test showed that the mean NDVI-570 score of the control plots significantly differed from the Temperature (p= 0.0024) and Drought (p= 0.0128) plots. For NDVI-509 these values were (p= 0.04) for control-drought and (p=0.006) for control-temperature The null-hypothesis that the difference between 2009 and 2013 for NDVI-570 and NDVI-509 had equal means was not rejected (p = 0.14), meaning that the NDVI showed no significantly differences between the 2009 and 2013 plots for both wavelengths.

16

(17)

Discussion:

 What is the effect of manipulating temperature and precipitation on the regrowth of heathland vegetation?

Cladonia and Rumex acetosella tend to remain relatively stable in control and temperature plots but show an increase in the drought plots. According to Ransijn (2014), drought tends to reduce the competitive ability of Calluna vulgaris. The regrowth of Calluna vulgaris is positively affected by an increase in temperature and negatively affected by an increase in droughts. Droughts could therefore positively affect the succession of other

competitive species such as Cladonia and Rumex acetosella. This could also explain why the cover of bare area shows the largest decrease in the drought plots (figure 9). Calluna vulgaris has a significant different growing pattern in the temperature plots (figure 11) compared to the control plots. According to Peñuelas et al . (2004) droughts alter the vegetative growth period and the flowering time span of common heathland species, but these are less affected by warming. Cladonia and Campylopus introflexus are practically not abundant in the 2013 plots which

implies that these species have relatively low growth rates after disturbances. Furthermore, this could be an indication that both can be considered as flexibility species. Hypnum jutlanicum decreases in all plots. According to the BLWG (2007) Hypnum benefits from the canopy cover of Calluna vulgaris which decreases incoming solar radiation. Regarding the predicted increase of Calluna vulgaris it is expected that the cover of Hypnum jutlandicum will benefit.

The species composition between 2015 and 2016 did not significantly change within a period of one year with an exception for the bare area in the drought plots (p = 0.035). A possible explanation related to this exception could be the defect of some drought plots in this specific year. All other species show differences but do not show extreme shifts in cover. These results imply that

appropriate time steps for monitoring vegetation dynamics in this area could still be sufficient if for 17

Figure 11: Growth of Calluna vulgaris (with standard deviation) in control, drought and temperature plots.

(18)

example the time steps are increased with one year. On the other hand according to Ransijn (2014) treatment effects generally appear after several years (>6 years), which emphasizes the importance of yearly monitoring in this research phase.

As showed in the results the cover of Calluna vulgaris in the temperature plots of 2009 was

significantly higher compared to all other groups. This implies that Calluna vulgaris benefits from the elevated temperature during the night-time warming. This result directly rejects the hypothesis that enhanced temperatures could increase microbial activity and thereby increases the nutrient

availability which is unflavoured by Calluna vulgaris. Apparently the increased temperature positively affects the succession of Calluna vulgaris. According to the results it is expected that increasing temperatures will lead to further expansion of the Calluna vulgaris. Droughts are likely to increase interspecific competition in the heathland ecosystem. As a result of the relative decrease of Calluna vulgaris other species such as Rumex acetosella, Cladonia and Dicranum scoparium could benefit.

The growth of Calluna vulgaris can partially be explained by the increased plant productivity. Figure 12 shows the NDVI plotted against the fraction of bare area. A Spearman correlation showed that the NDVI and fraction of Calluna vulgaris highly correlated (R= 0.72, p = 0.001). In this case the Spearman correlation is more appropriate compared to the Pearson correlation because the Pearson correlation measures the strength of two normally distributed

variables. In this particular case the fraction of Calluna vulgaris is not normally distributed (p = 0.0024). Regarding the fact that the Spearman correlation is higher than the Pearson correlation there is no direct linear relationship. Therefore a second degree function was used to fit the data. Figure 12 shows this correlation. It illustrates that an increase in Calluna vulgaris positively correlates with the NDVI , this effect weakens for higher cover levels. This indicates that enhanced temperatures have increased the NDVI (and therefore the plant 18

(19)

productivity) of Calluna vulgaris resulting in a larger increase of Calluna vulgaris in the temperature plots. Furthermore it notable that lower fractions of Calluna vulgaris, mainly from the drought plots, correspond to relative low NDVI values. This result strongly indicated that small changes in

precipitation and temperature affect the absorbance of light and thus could affect plant productivity. No significant differences were found between the 2009 and 2013 plots related to the NDVI, which indicates that plant productivity does not significantly change within this timescale, this seems to be surprising looking at the different fractions of bare in 2009 and 2013 (figure 7B). Apparently the correlation between NDVI and the fraction of bare area (R =0.4, p<0.1) can be considered as a moderately negative correlation. Other species show no clear relation with the NDVI measurements. Furthermore the effects of drought and temperature were approximately the same for 509 and 570 nm. This implies that the absorbance of different wavelengths reacts similar on small temperature and precipitation changes.

Further research:

To quantify the effects of the manipulation experiment it would be a valuable addition to measure the NDVI yearly . This way a time series is created that can be used to get a better understanding on how climate change affects the photosynthesis. Phenology is a key ecological process that involves plant growth, shifts in these phenophases are one the most accurate biological indicators of climate change (Walther, 2003; Jentsch et al. , 2009). NDVI shows a high seasonal variability throughout the year(Potter & Brooks,1998). Measuring this variability could be a powerful strategy to detect changes in the phenology of heathland species. Furthermore, it would be valuable to include soil

characteristics such as pH, CEC and N:P ratio to get an idea how a changing climate can influence soil characteristics. To better understand the impact of a changing climate , I would suggest to incorporate temperature and precipitation gradients by comparing obtained data from the Dutch experimental site with experiments in other climatological zones. Hereby, it is important that all experimental sites use the same methodology in their manipulation experiment.

19

(20)

Conclusion:

Small changes in precipitation and temperature are likely to affect the recovery of the heathland ecosystem. Species regeneration seems to be influenced by manipulating temperature and precipitation , especially the species composition of Calluna vulgaris tends to react to these

manipulations. Increased temperatures are likely to positively affect the recovery capacity of Calluna vulgaris, whether drought plots showed a reverse trend. The relative low increase of Calluna vulgaris in drought treatments offers chances for succession of other species such as Rumex acetosella, Dicranum scoparium and Cladonia. The increase of Calluna vulgaris in the warming plots is likely to correlate with the increased NDVI in these plots. The photosynthetic capacity of species significantly increases when comparing the control and warming plots. Apparently, small changes in climatological conditions can affect the succession of species after a disturbance. Whether these effects positively or negatively influence the stability of the ecosystem is not of primary concern. Fact is that these effects are commonly underestimated. This does underline the importance of better understanding these effects to maintain ecosystem stability in the future.

(21)

Acknowledgments:

- Albert Tietema - Henk Pieter Sterk

- Berend-Christiaan Wijers - Martijn Verhoog

- Jaleesa Schaap - Brand Timmer

- Artilerie Schiet Kamp (ASK) ‘t Harde

21

(22)

Literature:

Beier, C., Emmett, B., Gundersen, P., Tietema, A., Penuelas, J., Estiarte, M. & Williams, D. (2004). Novel

approaches to study climate change effects on terrestrial ecosystems in the field: drought and passive nighttime warming. Ecosystems, 7(6), 583-597.

BLWG.nl, (2007). Hypnum jutlandicum, Ecologie & verspreidng. Retreieved:

01-07-2016:https://www.verspreidingsatlas.nl/2792#.

Fagúndez, J. (2012). Heathlands confronting global change: drivers of biodiversity loss from past to future

scenarios. Annals of Botany, mcs257.

GeologievanNederland.nl. Retrieved: 22-04-2016 A

Grubb, P.J. (1998) A reassessment of the strategies of plants which cope with shortages of resources.

Perspectives in Plant Ecology, Evolution and Systematics, 1, 3-31.

IPCC, 2014: Climate Change 2014: Synthesis Report. Contribution of Working Groups I, II and III to the Fifth

Assessment Report of the Intergovernmental Panel on Climate Change [Core Writing Team, R.K. Pachauri and

L.A. Meyer (eds.)]. IPCC, Geneva, Switzerland, 151 pp.

Kopittke, G. R., Loon, E. van, E., Tietema, A., & Asscheman, D. (2013). Soil respiration on an aging managed

heathland: identifying an appropriate empirical model for predictive purposes. Biogeosciences, 10(5),

3007-3038.

Kröel-Dulay, G., Ransijn, J., Schmidt, I. K., Beier, C., De Angelis, P., De Dato, G., Kongstad, J. (2015). Increased

sensitivity to climate change in disturbed ecosystems. Nature communications, 6.

Latham, R.E. (2003). Shrubland longevity and rare plant species in the northeastern United States. Forest Ecology and Management, 185, 21- 39.

Mobaied S, Riera B, Lalanne A, Baguette M, Machon N. (2011). The use of diachronic spatial approaches and

predictive modelling to study the vegetation dynamics of a managed heathland. Biodiversity and Conservation

20: 73– 88.

Niinemets, Ü., Flexas, J. & Peñuelas, J. (2011) Evergreens favored by higher responsiveness to increased CO2. Trends in Ecology & Evolution, 26, 136-142.

NASA Earth Observatory, based on data provided by the University of Montana NTSG.

Peñuelas J, Gordon C, Llorens L, et al. (2004). Nonintrusive field experiments show different plant responses to

warming and drought among sites, seasons and species in a North–South European gradient. Ecosystems 2004;7:598-612.

Potter, C. S., & Brooks, V. (1998). Global analysis of empirical relations between annual climate and seasonality

of NDVI. International Journal of Remote Sensing, 19(15), 2921-2948.

Ransijn, Johannes. 2014. Changing Heathlands in a Changing Climate - Climate Change Effects on Heathland

Plant Communities. University of Copenhagen.

Schaap, J. (2015). The influence of climate change on heathland vegetation colonization and regeneration in Oldebroek, the Netherlands. University of Amsterdam. Amsterdam.

Walther, G.R.(2003). Plants in a warmer world. Perspectives in Plant Ecology, Evolution and Systematics ,6:169-185.

(23)

Appendix 1. Tables

Table 1a: Mean cover (%) of main species in different treatments (C = control, D = Drought and T = temperature)and year cut

S.D. Calluna Campylopus Cladonia Dicranum Hypnum Poaceae Polytrichum Rumex Bare

C2009 11.92 14.05 6.21 4.45 3.75 1.2 0.62 0 23.92 C2013 5.01 0 0 9.17 6.71 3.43 0.41 0 19.21 D2009 13.18 8.76 21.04 26.8 1.6 0.44 0.11 3.68 7.1 D2013 1.38 0 0.36 2.36 14.64 5.73 1.84 6.51 15.81 T2009 7.03 6.27 7.37 1.61 5.85 1.94 3.97 0.54 5.94 T2013 7.92 0 0 12.92 9.18 2.46 4.27 0.37 11.85

Table 1b: Standard deviation of mean cover (%) for main species in different treatments (C = control, D = Drought and T = temperature) and year cut

23

Knibbeler, S. (2016). ‘The impact of climate change on the regrowth of disturbed heathland’ University of Amsterdam. (%) Callun a Campylop us Clado nia Dicranu m Hypnu m Poacea e Polytrichu m Rume x Bare C2009 21.11 10.71 5.03 7.34 3.56 2.16 4.27 0 42.7 C2013 7.9 0 0 10.76 8.88 3.96 0.39 0 64.12 D2009 12.74 10.97 29.23 19 1.82 0.52 0.06 4.18 19.7 D2013 7.19 0 0.24 23.31 18.13 3.99 2.07 5.64 37.23 T2009 43.81 6.09 9.17 4.87 4.85 2.18 3.76 0.31 21.54 T2013 13.13 0 0 31 12.1 2.32 2.47 0.22 36.84

(24)

Table 2: p-values from multiple ANOVA provide information to which extent a category influences the cover (p<0.05: significant influence on cover).

Table 3 : Results of two sided T-test testing the null-hypothesis that the mean fraction of a species in a certain treatment does not differ between the dataset of 2015 and 2016. H = 1 if the null-hypothesis is rejected with a 5 % significance level.

24

p-value Treatment Year Interaction

Bare 0.02 0.03 0.94 Calluna 0.01 0 0.07 Campylopus 0.81 0.02 0.81 Cladonia 0.1 0.01 0.1 Dicranum 0.28 0.09 0.26 Hypnum 0.73 0.03 0.48 Poaceae 0.87 0.23 0.65 Polytrichium 0.39 0.39 0.17 Rumex 0.03 0.76 0.89 H p Bare_C 0 0.3253615 Bare_D 1 0.0353073 Bare_T 0 0.232705 Calluna_C 0 0.0584445 Calluna_D 0 0.1112854 Calluna_T 0 0.0502913 Campylopus_C 0 0.771751 Campylopus_D 0 0.4503989 Campylopus_T 0 0.5094551 Cladonia_C 0 0.7393262 Cladonia_D 0 0.8303471 Cladonia_T 0 0.2664687 Dicranum_C 0 0.9933547 Dicranum_D 0 0.1650324 Dicranum_T 0 0.3170755 Hypnum_C 0 0.700387 Hypnum_D 0 0.6910811 Hypnum_T 0 0.9541819 Poaceae_C 0 0.1395444 Poaceae_D 0 0.32888 Poaceae_T 0 0.1766858 Polytrichium_C 0 0.0838294 Polytrichium_D 0 0.4112109 Polytrichium_T 0 0.1909868

(25)

*NaN: not abundant

Table 4a: Mean change per year for main species calculated from a 5 year period .

25

(%)

Control Drought Temperature

Calluna 3.41 2.01 6.97 Campylopus 1.79 1.83 1.01 Cladonia 0.84 4.87 1.53 Dicranum 0.56 2.33 -0.91 Hypnum -1.74 -3.47 -1.97 Poaceae 0.19 -0.02 0.25 Polytrichium 0.71 -0.27 0.63 Rumex 0 0.36 -0.67 Bare -5.83 -7.61 -6.91

(S.d)

Control Drought Temperature

Calluna 2.82 3.37 2.27 Campylopus 1.68 2.68 2.89 Cladonia 1 3.43 4.28 Dicranum 4.15 3.4 6.12 Hypnum 2.26 2.1 2.89 Poaceae 1.04 1.26 0.89 Polytrichium 0.39 1.94 1.66 Rumex 0 2.36 1.4 Bare 6.97 4.06 6.28

(26)

Table 4b: Standard deviation of mean change per year

(%) Calluna Campylopus Cladonia Dicranum Hupnum Poaceae Polytrichum Rumex Bare

C1_2009 19.43 3.21 0.67 8.87 7.62 1.49 4.92 0 49.59 C2_2009 33.77 26.93 12.14 2.32 0.22 1.44 4.21 0 16.09 C3_2009 10.11 2 2.27 10.82 2.83 3.54 3.69 0 62.41 C1_2013 10.38 0 0 20.22 16.36 2.8 0.34 0 44.71 C2_2013 2.13 0 0 1.91 6.86 1.25 0 0 83.12 C3_2013 11.2 0 0 10.16 3.41 7.81 0.81 0 64.54 D1_2009 6.88 17.2 41.46 5.71 1.86 0.99 0.19 5.64 18.2 D2_2009 3.51 0.95 4.93 49.86 3.4 0.46 0 6.91 27.43 D3_2009 27.83 14.76 41.3 1.45 0.21 0.1 0 0 13.47 D1_2013 5.65 0 0.66 24.31 29.19 1.02 2.69 13.15 21.48 D2_2013 7.61 0 0.07 25.02 23.67 0.36 0 1.9 37.11 D3_2013 8.31 0 0 20.61 1.54 10.61 3.51 1.86 53.09 T1_2009 44.07 5.3 16.47 5.52 0.76 0 0 0 26.22 T2_2009 36.65 12.71 9.32 6.05 2.25 3.7 7.91 0 14.86 T3_2009 50.71 0.25 1.74 3.03 11.55 2.84 3.38 0.93 23.53 T1_2013 5.84 0 0 38.89 2.73 0.25 0 0 50.3 T2_2013 11.99 0 0 38.01 12.5 1.66 0 0 32.24 T3_2013 21.56 0 0 16.09 21.08 5.04 7.4 0.65 27.99 26

(27)

Table 5: Species cover(%) in all plots

(C = control, D = Drought and T = temperature)

Plot NDVI 570 NDVI 509 Plot NDVI 570 NDVI 509

C1-2009-1 0.6 0.71 D2-2013-1 0.45 0.6 C1-2009-2 0.52 0.66 D2-2013-2 0.46 0.58 C1-2009-3 0.47 0.57 D2-2013-3 0.45 0.61 C1-2009-4 0.43 0.54 D2-2013-4 0.41 0.61 C1-2013-1 0.56 0.71 D3-2009-1 0.58 0.68 C1-2013-2 0.46 0.58 D3-2009-2 0.4 0.53 C1-2013-3 0.52 0.66 D3-2009-3 0.39 0.51 C1-2013-4 0.51 0.64 D3-2009-4 0.56 0.7 C2-2009-1 0.6 0.75 D3-2013-1 0.42 0.55 C2-2009-2 0.58 0.7 D3-2013-2 0.44 0.55 C2-2009-3 0.58 0.74 D3-2013-3 0.44 0.55 C2-2009-4 0.59 0.76 D3-2013-4 0.52 0.65 C2-2013-1 0.48 0.59 T1-2009-1 0.68 0.8 C2-2013-2 0.46 0.58 T1-2009-2 0.56 0.66 27

(28)

C2-2013-3 0.4 0.5 T1-2009-3 0.56 0.68 C2-2013-4 0.38 0.47 T1-2009-4 0.48 0.56 C3-2009-1 0.47 0.58 T1-2013-1 0.51 0.64 C3-2009-2 0.48 0.61 T1-2013-2 0.55 0.66 C3-2009-3 0.48 0.6 T1-2013-3 0.5 0.63 C3-2009-4 0.5 0.64 T1-2013-4 0.49 0.62 C3-2013-1 0.53 0.66 T2-2009-1 0.57 0.71 C3-2013-2 0.58 0.72 T2-2009-2 0.57 0.71 C3-2013-3 0.6 0.74 T2-2009-3 0.53 0.67 C3-2013-4 0.55 0.69 T2-2009-4 0.63 0.78 D1-2009-1 0.46 0.58 T2-2013-1 0.56 0.71 D1-2009-2 0.43 0.56 T2-2013-2 0.6 0.75 D1-2009-3 0.45 0.58 T2-2013-3 0.59 0.75 D1-2009-4 0.55 0.67 T2-2013-4 0.57 0.7 D1-2013-1 0.49 0.65 T3-2009-1 0.55 0.69 D1-2013-2 0.47 0.61 T3-2009-2 0.49 0.61 D1-2013-3 0.56 0.67 T3-2009-3 0.59 0.73 D1-2013-4 0.5 0.63 T3-2009-4 0.63 0.79 D2-2009-1 0.46 0.57 T3-2013-1 0.57 0.71 D2-2009-2 0.47 0.6 T3-2013-2 0.63 0.76 D2-2009-3 0.47 0.6 T3-2013-3 0.65 0.78 D2-2009-4 0.45 0.58 T3-2013-4 0.59 0.72 28

(29)

Table 6: NDVI values in all plots

(C = control, D = Drought and T = temperature)

Appendix 2. Maps

29

(30)

(31)

31

(32)

(33)

33

(34)

(35)

35

(36)

(37)

37

(38)

(39)

39

(40)

(41)

41

(42)

(43)

43

(44)

(45)

45

(46)

(47)

The impact of climate change on the regrowth of disturbed heathland A manipulation experiment to identify changes in heathland species dynamics.