Analysing climatic drivers of
12
C/
13
C fractionation
during plant growth in a standardized phytometer
setup across Europe
BSc Thesis
Written by Julia van Heesewijk
Student Number 11033193
Under supervision of dr. A. Tietema
Amsterdam, 2th July 2018
University of Amsterdam
Institute of Interdisciplinary Studies
2 Figure title page: Schematic representation of an 13C isotope.
Adapted from: ‘Explainer: what is an isotope?’ by Koumoundouros, T. (n,d.). Retrieved from: https://theconversation.com/explainer-what-is-an-isotope-10688
3
Samenvatting
Dit onderzoek is gebaseerd op een grootschalig phytometer experiment dat heeft plaats gevonden langs een klimaat gradiënt in Europa. Aan het experiment, genaamd het Bayreuth Phytometer Initiatief, hebben verschillende universiteiten verspreid over tien Europese landen deelgenomen. Deze universiteiten zijn betrokken met ClimMani, een organisatie welke doelt op het verbeteren van de kwaliteit van klimaat gerelateerde manipulatie experimenten. Op elf verschillende locaties is een phytometer opgezet, waarin drie plantensoorten hebben gegroeid in zowel een standaard als een lokale bodem. Na een groeiperiode van 50-dagen zijn de planten gemeten en geoogst. De monsters die zijn verzameld tijdens dit experiment hebben uitgebreide mogelijkheden geboden voor onderzoek naar het effect van klimaat en bodem op plantengroei. In een onderzoek door van Dusseldorp (2018) is correlatie gevonden tussen verschillende klimatologische variabelen en 13C discriminatie in planten tijdens fotosynthese. Deze relatie is in lijn met de overheersende hypothese in de literatuur dat 13C discriminatie afneemt naarmate een plant meer droogte stress ervaart. Toch was de correlatie gevonden door van Dusseldorp (2018) kleiner dan verwacht. Van Dusseldorp (2018) raadde daarom een vervolg onderzoek aan, waarin een uitgebreidere klimaat gradiënt wordt bestudeerd. In dit onderzoek is de koolstof isotoop data van drie aanvullende locaties toegevoegd aan de bestaande dataset door van Dusseldorp (2018). Deze data is vervolgens gerelateerd aan drie verschillende klimatologische variabelen, namelijk de neerslag, de accumulatie van hitte in de bodem en een droogte index gebaseerd op de methode van Gaussen-Bagnouls. Voor alle drie de variabelen zijn relaties gevonden met de koolstof isotoop ratio’s die de bovengenoemde hypothese bevestigen. De sterkste correlatie is gevonden voor de Gaussen index, wat bevestigt dat deze variabele de klimaat gradiënt het beste weergeeft. Daarnaast heeft dit onderzoek bevestigd dat 13C discriminatie in planten beïnvloedbaar is op een korte tijdschaal. Zodoende kan worden bediscussieerd dat het meten van 13C discriminatie een geschikte methode is voor het onderzoeken van ecosysteem veranderingen met behulp van manipulatie experimenten.
Summary
The Bayreuth Phytometer Initiative is a phytometer experiment extended across a climate gradient in ten European countries. All participating universities in this experiment are connected to ClimMani, an organization aiming at improving the quality of climate manipulation experiments. At eleven European locations a phytometer was set up, in which three plant species were grown in a standardized and a local soil. After a 50-day growth period these samples were harvested and sent to the University of Amsterdam. The samples of this extended experiment created a variety of opportunities for research concerning the effects of climate and soil on different aspects of vegetation growth. In a research by van Dusseldorp (2018) correlation was found between isotope fractionation processes and climate. Van Dusseldorp (2018) recommended a follow up research with an extended climate gradient. This research has added carbon isotope fractionation data in plant material of three more locations to the existing data by van Dusseldorp (2018). This data has been related to three climatic variables which are the precipitation, the growing degree days and the Gaussen aridity index. Correlation has been found between carbon isotope fractionation for all three climatic variables. Thus confirming the hypotheses that carbon isotope fractionation decreases in more arid climatic conditions. The strongest correlation has been found for the Gaussen aridity index, which confirms that this variable provides the best representation for the climate gradient. Change in carbon isotope fractionation has been found over a 50-day growth period, it is therefore discussed that carbon isotope fractionation is sensitive to climatic changes on a short term. This indicates that carbon isotope fractionation is suitable as a common metric for ecosystem change.
4
Content
1. Introduction ... 5
1.1 Relevance ... 5
1.2 Carbon Isotope Fractionation ... 5
1.3 Climate Gradient ... 6
1.4 Climate Variables ... 8
1.5 Common Metrics for Ecosystem Change ... 9
1.6 Research Questions ... 9
2. Methods ... 10
2.1 Bayreuth Phytometer Initiative ... 10
2.2 Laboratory Analysis ... 11
2.3 Statistical Analysis and Data Management ... 12
3. Results... 13
3.1 Dead Biomass... 13
3.2 Gaussen Aridity Index ... 14
3.3 Growing Degree Days ... 16
3.4 Precipitation... 18
4. Discussion ... 21
4.1 Δδ13C ... 21
4.2 Biomass ... 21
4.3 Independent Variables ... 21
4.4 Local vs. Standard soil ... 22
4.5 Common Metrics ... 22 5. Conclusion ... 24 6. Acknowledgements... 25 7. References ... 26 7.1 Literature ... 26 7.2 Figures ... 27
5
1. Introduction
1.1 Relevance
Ecosystems and the services they provide are of great importance to humanity (Millennium Ecosystem Assessment, 2005). Climate change is expected to have a serious impact on the world’s ecosystems (Houghton, 2015). This has increased the need for research concerning the response of ecosystems to climate change (Cramer et al., 2001). Important methods for such research are manipulation experiments (De Boeck et al., 2015). Active within this field is a network of researchers named ‘Climate Change Manipulation Experiments in Terrestrial Ecosystems’ (ClimMani). The aim of this organization is to improve the quality of experiments and to facilitate the use of experimental data for climate change research (climmani.org).
Over a year ago Dr. P. Wilfahrt and Prof. Dr. A. Jentsch from the Bayreuth Center of Ecology and Environmental Research initiated a standardized phytometer experiment across Europe. Multiple Universities affiliated with ClimMani participated in this experiment, henceforth referred to as the Bayreuth Phytometer Initiative. All participants set up a phytometer, resulting in experimental sites on a climate gradient in ten different European countries. At all sites three types of common plant species were grown in a greenhouse. Thereafter, they were planted outdoor in as well standardized as local soils. At the end of a 50-day growth period the plants and the soil was harvested, measured and sent to the University of Amsterdam.
The samples of this extended experiment created a variety of opportunities for research concerning the effects of climate and soil on different aspects of vegetation growth. One of such researches was conducted half a year ago by Marleen van Dusseldorp. At the time a master student in the field of Earth Sciences at the University of Amsterdam. Van Dusseldorp (2018) dedicated her master’s thesis to ‘Disentangling climatic drivers of 15N and 13C plant data in a standardized phytometer setup’. In this master’s thesis, the samples of eight locations from the Bayreuth Phytometer Initiative were analysed. The samples originated from the following countries: Denmark (two locations), the Netherlands, Germany, Austria, Hungary, Serbia and Spain (van Dusseldorp, 2018).
The correlation found between the carbon isotopes and the climatic parameters by van Dusseldorp (2018) was weaker than expected. A follow up research, including data from more locations, was therefore recommended (van Dusseldorp, 2018). In this study the research by van Dusseldorp (2018) has been extended by adding the samples of the following countries to the existing data: Belgium, the Czech Republic and Estonia. These sites were chosen to result in a more complete climate gradient in comparison to van Dusseldorps (2018) research.
Where van Dusseldorp (2018) studied both the isotopes of nitrogen as carbon in plants and soil, in this research only the isotopes of carbon in plants have been studied. In the following paragraphs the theories underlying this research are explained.
1.2 Carbon Isotope Fractionation
Carbon has two stable isotopes which are 12C and 13C. The natural abundance of the lighter isotope is 98,9%, the natural abundance of 13C is 1,1% (Enoch et al., 1984). Research has concluded that plants discriminate against 13C while taking up CO2 in the process of photosynthesis (Farquhar et al, 1989, O’Leary, 1988). This is based on the finding that the abundance of 13C in plant tissue is lower in comparison to the abundance of 13C in the atmosphere (Farquhar et al., 1989). Farquhar et al., (1989) defines the cause of 12C/13C variation as fractionation effects.
The reason behind the occurrence of fractionation effects are distinctions in as well chemical as physical characteristics of the isotopes (O’Leary, 1988). Due to these distinctions, during some physical processes, such as photosynthesis, discrimination against heavier isotopes occurs. There is a clear distinction in the rate of 13C discrimination between C3 and C4 plants (O’Leary, 1988). 13C
6
Discrimination in C3 plants is relatively high (O’Leary, 1981). As mentioned earlier, in this research fractionation process will be analysed in three different plant species. As all three plant species have been characterized as C3 plants, this research will only focus on fractionation processes in the photosynthesis pathway of C3 plants.
Fractionation processes in C3 plants partly take place during the dissolvement and diffusion of CO2 in the internal gas space of plant cells (O’Leary, 1988). However, most of the fractionation effects occur during the carboxylation of the enzyme ribulose bisphosphate (O’Leary, 1988). While experiencing drought stress, plants close their stomata, which results in a lower 13C discrimination (Farquhar et al., 1989). Therefore, higher 13C abundances are found in plant biomass originating from more arid climates (Klaus et al., 2016, Stewart et al., 1995).
In this research isotope fractionation is expressed in δ13C, formula 1 describes how δ13C has been calculated.
δ13C = [ 𝑅(𝑠𝑎𝑚𝑝𝑙𝑒)
𝑅(𝑠𝑡𝑎𝑛𝑑𝑎𝑟𝑑)− 1] × 100 𝑖𝑛 ‰ 𝑤𝑖𝑡ℎ
𝑅 = 13𝐶𝑂2 / 𝐶𝑂12 2
Formula 1: Calculation of δ13C, method retrieved from O’Leary (1988)
In formula 1, R(standard) represents a widely used standard 13C/12C ratio in stable isotope geochemistry. This standard is based on a limestone formation in South Carolina (O’Leary, 1988). In comparison to this standard, the ratio of 13C/12C is smaller in the plant samples as a result of 13C discrimination. Therefore, the δ13C values in this research are negative.
1.3 Climate Gradient
The hypotheses underlying this research is that less negative values of δ13C are found in plants grown in more arid climates (Klaus et al., 2016, Stewart et al., 1995). This hypotheses will be researched by comparing the δ13C values in plant material obtained from different locations on a climate gradient in Europe. The locations from which this plant material originates can be seen on the maps in figure 1.
7
Figure 1: Left: Sampling locations analysed by van Dusseldorp (2018). Right: All sampling locations analysed in this research, locations additional to van Dusseldorp (2018) marked in blue.
As mentioned earlier, samples originating from eight sites had already been analysed by van Dusseldorp (2018). For this research, the data from three additional sites has been added to the existing data. These three locations are marked by the blue pinpoints in figure 1.
Figure 2 visualizes the climate gradient studied in this research by comparing the mean annual temperature in °C to the mean annual precipitation in mm for every location.
Figure 2: Visualisation of the climate gradient by comparing mean annual temperature and mean annual precipitation.
8
1.4 Climate Variables
This research has related δ13C values in plant material to three climatic variables which are all based on measurements obtained at the sampling sites. The three independent variables are the precipitation on site, the growing degree days (GDD) and the Gaussen aridity index. The first variable has been directly measured at the sites during the growth period, the latter two variables were calculated later. Figure 3 visualises the sample locations related to the precipitation and soil temperature measured at the sites.
Figure 3: Visualisation of the climate gradient by comparing the average soil temperature and precipitation measured on site over the growth period.
The GDD is indicative for the accumulation of heat in the soil over the 50-day growth period. While calculating GDD, maximum and minimum temperatures are not averaged out. Therefore, GDD is a more reliable variable in comparison to the average temperature (Miller et al., 2001). In this research, the GDD is based on the daily maximum and minimum soil temperatures and the average daily soil temperature. A sine wave was fitted to the minimum and maximum temperatures and the area under the curve was summed for all days.
The Gaussen aridity index is based on the Gaussen-Bagnouls classification method. Formula 2 represents the calculation for the Gaussen aridity index used in this research. The Gaussen-Bagnouls method is originally applied to make distinctions among dry and wet months (Nikolova & Mochurova, 2012). The demarcations used in this method for climatic classification are given along with formula 2. An advantage of the use of the Gaussen aridity index is that it combines temperate and precipitation and therefore provides a more complete representation of the climate of a location. For this reason, combining temperature and precipitation into one variable is a popular method in vegetation studies (Baltas, 2007). In this research the Gaussen index has been applied to create a climate gradient of the sample locations based on aridity.
9 𝐺𝑎𝑢𝑠𝑠𝑒𝑛 𝑎𝑟𝑑𝑖𝑡𝑦 𝑖𝑛𝑑𝑒𝑥 = 𝑃𝑟𝑒𝑐𝑖𝑝𝑖𝑡𝑎𝑡𝑖𝑜𝑛 50𝑑𝑎𝑦 𝑔𝑟𝑜𝑤𝑡ℎ 𝑝𝑒𝑟𝑖𝑜𝑑 𝐴𝑣𝑒𝑟𝑎𝑔𝑒 𝑑𝑎𝑖𝑙𝑦 𝑠𝑜𝑖𝑙 𝑡𝑒𝑚𝑝𝑒𝑟𝑎𝑡𝑢𝑟𝑒 𝑤𝑖𝑡ℎ > 3 𝐻𝑢𝑚𝑖𝑑 3 − 2 𝑆𝑒𝑚𝑖 𝐻𝑢𝑚𝑖𝑑 < 2 𝐴𝑟𝑖𝑑
Formula 2: Calculation for the Gaussen aridity index and demarcations for classifying climates. Demarcations based on Gabriels (2007).
1.5 Common Metrics for Ecosystem Change
As has been elaborated on in the previous paragraphs, this research partly aims at relating carbon isotope fractionation processes to climate variables. However, there is a more extensive aim underlying this research. As has been mentioned earlier, the aim of ClimMani is to improve the quality of manipulation experiments in the field of ecosystem response research. In line with this objective, this research aims at discussing whether 13C discrimination is suitable as a common metric for ecosystem change. The definition for common metrics of ecosystem change as interpreted in this research is based on 5 requirements stated by unpublished literature by Tietema (2017).
Common metrics must be indicative for ecosystem functioning, they are required to be sensitive enough to show significant changes with the time scale of an average manipulation experiment, they must be interpretable, they require (almost) non-destructive sampling for analysis and they are required to be relatively cheap and easily accessible for researchers (Tietema, 2017).
This research investigates if 13C fractionation is sensitive to climatic change on a short term, with the aim of discussing whether 13C fractionation is suitable as a common metric for ecosystem change. Tietema (2017) and van Dusseldorp (2018) have already discussed that carbon isotope fractionation is a suitable method for ecosystem change. This research will pursue this discussion by investigating if climatic variables influence isotope fractionation during a 50-day growth period.
1.6 Research Questions
Based on the theories and hypotheses elaborated in the previous paragraphs, this research aims at answering the following research questions:
How is 12C/13C fractionation in plants related to climate variables during a fifty-day growth period?
Does this make 13C discrimination suitable as a common metric for ecosystem change?’
These questions have been answered by analysing the change in δ13C values plants after a 50-day growth period in standardized phytometer setups on a European climate gradient. The δ13C has been related to the precipitation of the 50-day growth period, the GDD and the Gaussen aridity index. Based on these comparisons it has been discussed whether 13C discrimination is a suitable common metric for ecosystem change.
10
2. Methods
2.1 Bayreuth Phytometer Initiative
This research is based on samples originating from the Bayreuth Phytometer Initiative. This section describes how the Bayreuth phytometer experiment has been conducted.
All participants of the Bayreuth Phytometer Initiative followed a protocol by Wilfahrt et al., (2017). The complete protocol can be found in Appendix I. All participant received the same materials and plant seeds to set up the phytometer. The three plant species chosen for the phytometer experiment were Trifolium Pratense, Dactylis Glomerata and Plantago Lanceolate (figure 4). All three plant species are common species throughout Europe.
Figure 4: Plant species grown in the Bayreuth phytometer experiment.
Reprinted from ‘Wilde Planten in Nederland en België’. By Dijkstra, K. & Riemsma, W. (n,d). Retrieved from: https://wilde-planten.nl
In the first stage of the phytometer experiment the plants were germinated in greenhouses at all locations for 7 weeks. Thereafter, the plants were planted outdoors in ten pots, of which five were filled with local soil and five with a standardized soil. The standard soil consisted of vermiculite with osmocote fertilizer. The seedlings were planted, along with some measuring devices, according to the scheme represented at the left of figure 5. The pictures in figure 5 were taken at the phytometer setup of the University of Amsterdam, located in Oldebroek. The 50-day growth periods were based on the peak in the site specific growing season. For Oldebroek this period lasted from the 3th of July to the 22th of august 2017. Trifolium Pratense Red Clover Dactylis Glomerata Cocksfoot Plantago Lanceolata Ribwort Plantain
11
Figure 5: Left: Scheme used to plant seedlings and measuring devices. Reprinted from Wilfahrt et al., (2017). Right: Pictures of the phytometer setup in Oldebroek, sampling locations The Netherlands.
The device called the TidbiT datalogger was planted in the pots in order to measure the soil temperature. This data has been used to determine the GDD and the Gaussen aridity index. As the temperature was measured in both local and standard soil the two variables are different for both soil types. The other devices planted in the pots are irrelevant to this research. At all sites the precipitation was measured for the 50-day growth period.
After the 50-day growth period all plants were harvested. Subsequently, the biomass of both the dead and the alive plant biomass was measured. The plant species in each pot were separately sampled, resulting in 30 samples per location. All samples were then sent to de University of Amsterdam.
2.2 Laboratory Analysis
The techniques used in the laboratory for this research are exactly the same as the techniques used by van Dusseldorp (2018). The δ13C in the plant samples was determined by using EA-IRMS, which combines element analyses with isotope ratio mass spectrometry. Before being able to carry out this analysis the samples were pre-processed.
All samples have been milled in a milling machine used for pre-processing plant material (figure 6a). During this process dead and alive material was blended. The distinction in δ13C in dead and alive material was considered not to be relevant to the research. In addition, the quantity of dead material of many samples was not enough to analyse.
Subsequently, all milled samples were folded in triplicated into a tin capsule (figure 6b&d). Each capsule contained 5-10mg of sample material. In addition to the samples, a standard consisting of 1-5mg acetanilide was folded and weighted 60 times (figure 6c&d). This is a standard procedure while using EA-IRMS. The acetanilide capsules are used to calibrate the machine. In order to measure carbon isotopes in the EA-IRMS the samples have to be transposed into gas phase, meaning the samples are combusted in order for carbon to form CO2.
12
Figure 6: A: Milling machine used to pre-process the plant samples. B: Tin boat containing sample. C: Tin boat containing acetanilide. D: Tray containing capsules.
2.3 Statistical Analysis and Data Management
The statistical analysis of the data for this research has been conducted in MATLAB. The δ13C is the dependent variable in this research. The independent variables are precipitation, the GDD and the Gaussen aridity index. All independent variables have been tested for correlation with the dependent variable by means of regression analysis. This analysis has been substantiated with different means of visualisation.
As has been explained earlier, all locations received the same plant seeds, and in the first phase of the phytometer experiment these seeds were germinated in greenhouses at the different locations. The initial purpose of this greenhouse period was to germinate the plants under uniform environmental conditions at all locations. However, according to van Dusseldorp (2018) the greenhouse conditions were likely divergent at different locations. This has been based on the finding that the δ13C values in the plant material after the greenhouse period deviated for the different locations (van Dusseldorp, 2018). Formula 3 represents the correction made to overcome these differences.
Δδ13C = δ13Cℎ𝑎𝑟𝑣𝑒𝑠𝑡− δ13C𝑔𝑟𝑒𝑒𝑛ℎ𝑜𝑢𝑠𝑒 𝑖𝑛 ‰
Formula 3: Correction for deviation in greenhouse conditions. Method based on van Dusseldorp (2018).
As can be seen in formula 3, the δ13C values in the plant material after the greenhouse phase are subtracted from the δ13C values in the plant material after harvest. This correction is necessary as for this research the change in δ13C during the 50-day growth period is of relevance. In other words, only the change in δ13C after the greenhouse period is relevant. Therefore, in this research Δδ13C, as calculated in formula 3, has been analysed as the dependent variable.
13
3. Results
3.1 Dead Biomass
Figure 7 visualizes the Δδ13C related to the Gaussen aridity index for 10 locations. In both van Dusseldorp (2018) as in this research the Δδ13C values for Spain were lower than expected. This can be seen in figure 7. Note that this figure only represents an example of the results including the location Spain, nevertheless the other results showed a similar deviation in the Δδ13C values for Spain.
Figure 7: Average Δδ13C values per plant species, related to the Gaussen aridity index. Based on averages of local and standard soil.
The weight of dead plant biomass has been related to the weight of the total plant biomass. This resulted in the percentages of dead biomass for all sites. The percentages are based on the weighted plant biomass in mg after harvest.
Trifolium (%) Dactylis (%) Plantago (%) Total (%) AW 2.52 1.65 1.82 1.99 TC 1.42 15.14 1.84 3.41 DO 12.38 21.47 1.99 11.59 BC 80.81 8.40 9.21 12.68 SE 0 4.52 0.44 1.03 MC 26.85 6.14 2.56 12.50 NE 0 0 0 0 GA 86.92 59.19 87.82 74.35 BT 16.53 8.10 4.70 13.20 VI 0.48 0 0 0.06 KI 80.09 24.38 9.64 18.40 Total 9.62 10.74 4.52 8.30
Table 1: Percentages of the weight of dead plant biomass compared to the total weight of the plant biomass for all locations. Based on average values in local soil and standard soil.
-2 0 2 4 6 8 10 Δδ 13C
Gaussen aridity index
Δδ
13C compared to Gaussen Index
14
The percentages of dead biomass per plant species and in total have been tested for outliers. The boxplots of this analysis are given in figure 8.
Figure 8: Boxplots of the percentages of dead plant biomass compared to total plant biomass biomass in all three plant species and of total of the plant species.
As can be concluded from figure 8, the percentage of dead biomass in Spain is an outlier compared to the percentages of dead biomass of the other locations. Except for the percentage of dead biomass of Trifolium Pratense, for this location the percentage of dead biomass of Spain is the upper adjacent value. Figure 8 also shows that the percentages of dead biomass for Trifolium are relatively higher compared to the other three species.
Based on the above described findings, the location Spain has been removed from the dataset. The data from the Spain location is not included in the continuation of the results.
3.2 Gaussen Aridity Index
This research has related Δδ13C as a dependent variable to the Gaussen aridity index. The results of this comparison are visualized in this paragraph. Note that the value of Denmark II is missing in the results of this data. The reason for this is that a climate variable on which the Gaussen index is based is unknown for this location.
The graphs in figure 9 visualize the Δδ13C values related to the Gaussen aridity index. There is a negative trend in Δδ13C values in plant material grown in both local as standard soil. Figure 9 also shows that the Δδ13C values for the species Plantago are negative at some locations.
15
Figure 9: Average Δδ13C values per plant species related to the Gaussen aridity index. Left: Plants grown in standard soil. Right: Plants grown in local soil.
Figure 10 represents a scatterplot in which the values per soil and per species of Δδ13C and the Gaussen aridity index are visualized. A first order polynomial is fitted through the scatterplot to visualize the linear trend within the data. This trend is negative for all plant species growing in both soils.
Figure 10: Scatterplot of the values of Δδ13C per plant species per site, related to the Gaussen aridity index. With a first order polynomial fitted to visualise the regression lines of the data. Purple arrows indicate that the labels on the x-axis are stretches as values lie close together.
16
Table 2 gives the correlation coefficients and the p-values for the Δδ13C and the Gaussen aridity index. For the Gaussen aridity index, there is a weak to strong negative correlation with Δδ13C. This correlation is significant for all species in both soil types.
Table 2: Correlation coefficients Δδ13C and Gaussen aridity index. Green markings markings indicate a negative correlation. Blue markings indicate
significant p-values.
3.3 Growing Degree Days
The second independent variable to which Δδ13C has been related is the GDD. This comparison is visualized in figure 11. This figure does not show a clear trend.
Figure 11: Average Δδ13C values per plant species related to the GDD. Left: Plants grown in standard soil. Right: Plants grown in local soil.
The Δδ13C values per plant species per soil have been scattered against the GDD in figure 12. A first order polynomial has been fitted through these points to indicate linear relationships. This figure shows that there is a positive linear relation between the data points. However, the Estonian values do not align with this linear trend.
Δδ13C – Gaussen LS SS TP r-coefficient -0.5609 -0.4319 p-value 1.12*10-4 0.0048 DG r-coefficient -0.6947 -0.7338 p-value 8.5*10-8 2.167*10-8 PL r-coefficient -0.6446 -0.6151 p-value 1.34*10-6 1.459*10-5 Total Per soil r-coefficient -0.6254 -0.5754 p-value 6.61136*10-16 1.8140*10-12 Total r-coefficient -0.5936 p-value 1.46*10-26
17
Figure 12: Scatterplot of the values of Δδ13C per plant species per site, related to the GDD. With a first order polynomial fitted to visualise the regression lines of the data.
Table 3 gives the correlation coeffects and the p-values for Δδ13C and the GDD. There is moderate significant correlation between the two variables for Dactylis and Plantago in both soils. There is no correlation between Δδ13C and GDD for Trifolium neither of the soil types.
Δδ13C – GDD LS SS TP r-coefficient 0.2166 -0.0644 p-value 0.1392 0.6670 DG r-coefficient 0.6030 0.6695 p-value 22,4*10-6 1.44*10-7 PL r-coefficient 0.5250 0.5093 p-value 6.46*10-5 2.186*10-4 Total Per soil r-coefficient 0.4545 0.3809 p-value 4.0471*10-9 2.4813*10-6 Total r-coefficient 0.4216 p-value 3.51*1014
Table 3: Correlation coefficients Δδ13C and GDD. Orange markings indicating positive correlation. Blue markings indicating significant p-values.
18
3.4 Precipitation
The last independent variable to which Δδ13C has been related is precipitation. This comparison is visualized in figure 13. In this figure a negative trend can be distinguished.
Figure 13: Average Δδ13C values per plant species related to the precipitation over the 50-day growth period. Left: Plants grown in standard soil. Right: Plants grown in local soil
The upper two graphs of figure 14 represent the scatterplot of Δδ13C and precipitation for 10 locations. The first order polynomial fitted through this trend indicates a negative relation between the variables. The x-axis shows that the precipitation values of the locations are distributed unevenly. Especially the precipitation measured in Austria is far higher compared to the other precipitation values. In order to visualize the results clearer, the bottom two graphs of figure 14 zoom in on the above figures, leaving Austria out.
19
Figure 14: Scatterplot of the values of Δδ13C per plant species per site, related to the precipitation over the 50-day growth period. With a first order polynomial fitted to visualise the regression lines of the data. Purple arrows indicate that the labels on the x-axis are stretches as values lie close together.
20
Table 4 gives the correlation coefficients for Δδ13C and precipitation. All coefficients indicate a significant weak negative correlation for Δδ13C and precipitation.
Δδ13C – Precipitation LS SS TP r-coefficient -0.4102 -0.3569 p-value 0.0038 0.0138 DG r-coefficient -0.4506 -0.4788 p-value 8.03*10-4 5.015*10-4 PL r-coefficient -0.3941 0.4614 p-value 0.0038 5.798*10-4 Total Per soil r-coefficient -0.4088 -0.3829 p-value 1.7062*10-7 2.1679*10-6 Total r-coefficient -0.3944 p-value 1.86*10-12
Table 4: Correlation coeffients Δδ13C and Precipitation. Green markings indicate positive correlation. Blue markings indicate significant
p-values.
21
4. Discussion
4.1 Δδ
13C
According to van Dusseldorp (2018) differences in greenhouse conditions among the locations led to deviating δ13C values after the greenhouse period. For this reason, Δδ13C has been used as the dependent variable in this research. Δδ13C is calculated by means of formula 3. As a result of drought stress, 13C abundance was expected to increase during the 50-day growth period, in other words, Δδ13C values in this research were expected to be positive.
As mentioned in the results, negative values of Δδ13C were observed at some locations for the species Plantago. These locations are Denmark I, Denmark II and Germany. This indicates that the 13C uptake of plants has decreased at these locations after the greenhouse period as a result of fractionation processes. Van Dusseldorp (2018) excluded that a difference in biomass can have caused these negative values. However, the δ13C values for these sites were already deviating after the greenhouse period (van Dusseldorp, 2018). Van Dusseldorp (2018) therefore suggests these negative Δδ13C values are caused by the initial differences at the greenhouses of these locations. Increased atmospheric CO2 alleviates drought stress in plants (Field, Jackson & Mooney, 1995, Wall, 2001). Higher CO2 abundance during the greenhouse period can have resulted in less drought stress and therefore more 13C discrimination. If the effect of alleviated drought stress continued after the greenhouse period this can explain the negative Δδ13C of Denmark I, Denmark II and Germany.
4.2 Biomass
As described in paragraph 3.1 the Δδ13C values in plant material from Spain were lower than expected. High percentages of dead plant biomass related to total plant biomass were found for Spain. For two out of three plant species, these percentages are outliers compared to the percentages of dead biomass in other locations. Spain has therefore been removed from the dataset. It is likely that the deviating Δδ13C values of Spanish samples are related to the high percentages of dead biomass. The plants at the Spanish phytometer most likely suffered from severe drought stress and mostly died. Therefore the measured Δδ13C values were not as high as expected.
As can be seen in figure 8, the percentages of dead biomass for the species Trifolium are generally higher compared to the other two species. The percentage of dead biomass for Spain is not an outlier for Trifolium, it is the upper adjacent value. Based on these boxplots it can be concluded that more dead biomass was harvested for Trifolium, indicating that this species suffered the most from drought stress compared to the other two species.
4.3 Independent Variables
For the GDD a positive relation was found with Δδ13C. A higher GDD indicates a higher heat accumulation in the soil and thus more drought stress and a higher 13C abundance. For the precipitation a weak negative relation has been found with Δδ13C. More precipitation leads to less drought stress and therefore less 13C. A negative relation was found between Δδ13C and the Gaussen aridity index. Meaning that plant material from locations with lower indices had higher Δδ13C values. All three relations as described above are in line with the hypotheses.
The correlation coefficients indicated the strongest correlation between the Gaussen aridity index and Δδ13C. By comparing the scatterplots of all three independent variables it can be concluded that the Δδ13C values are the least deviating whilst related to the Gaussen aridity index. This confirms that the climate gradient based on the Gaussen aridity index is most fitted to relate to the Δδ13C. As the Gaussen aridity index combines temperature and precipitation and therefore provides a better representation of climate.
22
To elaborate, figure 11 and 12 shows that the Δδ13C values do not fit the positive linear trend between Δδ13C and the GDD. The Estonian Δδ13C values are higher than expected. In addition, figure 13 and 14 show that the Estonian Δδ13C values compared to the precipitation are lower than expected. The precipitation measured in Estonia is similar to the precipitation measured in Serbia. However, the Δδ13C values for Estonia are lower than the Δδ13C for Serbia. The deviations of Estonia for precipitation and GDD can be explained by looking at figure 3. This figure shows that at the Estonian locations the lowest precipitation and the lowest soil temperature was measured. When only taking into account one of these variables, and thus placing Estonia for both variables at the far left of the climate gradient, the Δδ13C are either higher or lower than expected. This example confirms that the Gaussen aridity index, which combines temperature and precipitation, gives a better representation of the climate gradient in comparison to precipitation and GDD.
The correlation coefficients in table 3 show that no correlation has been found for GDD and Δδ13C for the species Trifolium. As mentioned earlier, the high percentages of dead biomass of the species Trifolium indicates that the species Trifolium suffered more from drought stress compared to the other species. Figure 12 shows the deviating values of the Estonian location as discussed above. It also shows the Estonian deviations from the trend line are highest for the specie Trifolium. According to the Gaussen aridity index (figure 10) Estonia is classified as the second or third driest location. It can therefore be explained that the highest Δδ13C values in Estonia are found for Trifolium, as it has been discussed that this specie likely suffered most from drought stress. For these reasons it is argued that the lack of correlation between Δδ13C and the GDD for the species Trifolium is likely caused by these deviating values of Estonia.
4.4 Local vs. Standard soil
For all variables and all plant species similar correlation has been found with Δδ13C in both local and standard soil. Therefore, no noteworthy differences have been found between Δδ13C in plants grown on standard soil in comparison to plants grown on local soils. For the GDD and the Gaussen aridity index there is a difference between the order of the locations within the climate gradients used in the figures. This is caused by differences in the measured soil temperatures of local soils in comparison to standard soils. As both the GDD and the Gaussen aridity index are based on soil temperature this results in different climate gradients for local and standard soil.
4.5 Common Metrics
As mentioned earlier, according to Tietema (2017) common metrics must be indicative for ecosystem functioning, they are required to be sensitive enough to show significant changes with the time scale of an average manipulation experiment, they must be interpretable, they require (almost) non-destructive sampling for analysis and they are required to be relatively cheap and easily accessible for researchers. According to Chaves, Maroco & Pereira (2003), researching stable carbon isotopes is a powerful tool to study plant responses to the environment ranging from the cellular to the ecosystem level. To elaborate, 13C discrimination is frequently used to explain the process of photosynthesis and environmental influences on this process (Farquhar et al., 1989). It can therefore be concluded that 13C discrimination is indicative for ecosystem change and meets the first requirement for common metrics. Measuring 13C abundance in plant material is a widely used method, for example 13C is commonly used to investigate CO2 fixation pathways and CO2 efficiency in plants (O’Leary, 1988). This confirms that 13C discrimination as a common metric is interpretable, relatively cheap and easily accessible for researchers.
This research has investigated whether 13C discrimination is sensitive to short term changes. The change in Δδ13C values over a 50-day period have been related to three climatic variables. As all
23
three variables are generally correlated to Δδ13C it can be discussed that the second requirement for common metrics is met.
24
5. Conclusion
The first research question that has been answered in this research is: How is 12C/13C fractionation in plants related to climate variables during a fifty-day growth period?
12C/13C Fractionation has been translated into Δδ13C, with higher Δδ13C indicating more 13C and less 12C/13C fractionation. There is a significant positive relation between Δδ13C and the GDD. Meaning that less fractionation occurred in plants grown in soils with a higher GDD. For the precipitation a weak significant negative relation has been found with Δδ13C. Meaning that more fractionation occurred in plants at locations with more precipitation. The strongest correlation has been found for the Gaussen aridity index and Δδ13C. This relation is negative, meaning that plant material from location with lower indices had higher Δδ13C and thus lower 12C/13C fractionation occurred. It has also been discussed that the Gaussen aridity index has provided at more representative climate gradient for this research. The second research question that has been answered is: Does this make 13C discrimination suitable as a common metric for ecosystem change? Δδ13C showed significant changes over the 50-day growth period. Therefore, the second requirement for common metrics for ecosystem change as stated by Tietema (2017) is met. Further, it has been discussed that 13C discrimination meets the additional requirements for common metrics of ecosystem change. It is therefore concluded that 13C discrimination is suitable as a common metric for ecosystem change.
25
6. Acknowledgements
Firstly, I want to express my sincere thanks to Dr. A. Tietema, who supervised this BSc project. I want to thank Dr. Tietema for providing the research subject of this thesis and all associating contacts and facilities. In addition, I want to thank Dr. Tietema for the guidance he provided and the expertise he shared in the process of writing this thesis.
Secondly, I want to thank M. van Dusseldorp, as her MSc thesis has been a fundamental guide while writing this BSc thesis. Van Dusseldorp also shared valuable skills in the laboratory, which enabled me to start the work in the laboratory uncomplicatedly.
I also want to thank the staff in the laboratory who were always available to answer questions regarding the laboratory work. A special thanks goes out to Dr. E. de Rijke who guided me patiently during the process of folding and weighting the samples. I am also very grateful for the help of J. Schoorl, who carried out the EA-IRMS analysis and provided not only an organized dataset but also extensive explanation on the analysis.
Finally, the most important fundament of this research is the Bayreuth phytometer experiment. I would therefore like to thank Dr. P. Wilfahrt and Prof. Dr. A. Jentsch for initiating this valuable experiment. In addition, this experiment would not have been possible without the ClimMani members who participated in the experiment.
26
7. References
7.1 Literature
Baltas, E. (2007). Spatial distribution of climatic indices in northern Greece. Meteorol. Appl. 14: 69–78 (2007). DOI: 10.1002/met.7
Chaves, M. M., Maroco, J. P. & Pereira, J. S. (2003). Understanding plant responses to drought – from genes to the whole plant. Functional Plant Biology, 2003, 30, 239-264.
De Boeck, H. J., Vicca, S., Roy, J., Nijs, I., Milcu, A., Kreyling, J., … Beier, C. (2015). Global Change Experiments: Challenges and Opportunities. BioScience, 65(9), 922–931.
https://doi.org/10.1093/biosci/biv099
Enoch, H. Z., Carmi, I., Rounick, J. S., & Magaritz, M. (1984). Use of carbon isotopes to estimate incorporation of added CO2 by greenhouse-grown tomato plants. Plant physiology, 76(4), 1083-1085.
Farquhar, G.D., Ehleringer, J.R., Hubick, K.T. (1989). Carbon Isotope Discrimination and Photosynthesis. Annual Review of Plant Physiology and Plant Molecular Biology, 40, 503-537. https://doi.org/ 10.1146/annurev.arplant.40.1.503
Field, C. B., Jackson, R. B., Mooney, H. A. (1995). Stomatal responses to increased CO2: implications from the plant to the global scale. Plant, Cell and Environment (1995) 18, 1214-1225.
Gabriels, D. (2007). Aridity and drought indices. Retrieved June 27, 2018, from http://indico.ictp.it/event/a06222/material/4/2.pdf
Hougthon, J. (2015). Global Warming: The Complete Briefing (5th ed.). Cambridge, United Kingdom: Cambridge University Press.
Klaus, V. H., Hölzel, N., Prati, D., Schmitt, B., Schöning, I., Schrumpf, M., … Kleinebecker, T.
(2016). Plant diversity moderates drought stress in grasslands: Implications from a large real-world study on 13 C natural abundances. Science of the Total Environment, 566, 215–222. McMaster, G. S., Wilhelm, W. W. (1997). Growing degree-days: one equation, two interpretations. Agricultural and Forest Meteorology 87 (1997) 291-300. https://doi.org/10.1016/S0168- 1923(97)00027-0
Millennium Ecosystem Assessment (2005). Ecosystems and Human Well-Being: Biodiversity Synthesis. World Resources Institute, Washington, D.C. (USA). Retrieved from:
https://www.millenniumassessment.org/documents/document.354.aspx.pdf Miller, P., Lanier, W., & Brandt, S. (2001). Using growing degree days to predict plant stages.
Ag/Extension Communications Coordinator, Communications Services, Montana State University-Bozeman, Bozeman, MO.
27
Nikoloca, N., Mochurova, M. (2012) Changes in Air Temperature and Precipitation and Impact on Agriculture. Forum geografic. Studii şi cercetări de geografie şi protecţia mediului Volume XI, Issue 1 (June 2012), pp. 81-89 (9) http://dx.doi.org/10.5775/fg.2067-4635.2012.044.i O'Leary, M. H. (1981). Carbon isotope fractionation in plants. Phytochemistry, 20(4), 553-567. O’Leary, M., H. (1988) Carbon Isotopes in Photosynthesis. BioScience, Vol. 38, No. 5 (May, 1988), pp. 328-336. Retrieved from: http://www.jstor.org/stable/1310735
Stewart, G. R., Turnbull, M. H., Schmidt, S., & Erskine, P. D. (1995). 13C natural abundance in plant communities along a rainfall gradient: a biological integrator of water availability. Functional
Plant Biology, 22(1), 51-55.
Tietema, A. (2017). Urgent need for common metrics to compare the results of climate manipulation experiments in soil-plant systems: window of applicability of 13C and 15N natural
abundance? Research proposal, University of Amsterdam.
Van Dusseldrop, M. (2018). Disentangling climatic drivers of 15N and 13C plant data in a standardized phytometer setup (Master’s thesis, University of Amsterdam).
Wall, G. W. (2001). Elevated atmospheric CO2 alleviates drought stress in wheat. Agriculture, Ecosystems and Environment 87 (2001) 261–271
Wallén, C. C. (1967). Aridity Definitions and Their Applicability. Geografiska Annaler. Series A, Physical Geography, Vol. 49, No. 2/4, Landscape and Processes: Essays in Geomorphology (1967), pp. 367-384. Retrieved from: http://www.jstor.org/stable/520903
Wilfahrt, P., Berauer, B., von Hessberg, A., Jentsch, A. (2017). The Bayreuth Phytometer - A common metric for community ecology.
7.2 Figures
Koumoundouros, T. (n,d.). Explainer: What is an isotope? Retrieved from: https://theconversation.com/explainer-what-is-an-isotope-10688
Dijkstra, K. & Riemsma, W. (n,d). Wilde Planten In Nederland en België. Retrieved from: https://wilde- planten.nl
28
Appendix I: Protocol Bayreuth Phytometer Experiment
The Bayreuth Phytometer – A common metric for community ecology
Peter Wilfahrt, Bernd Berauer, Andreas von Hessberg, Anke Jentsch
Disturbance Ecology, BayCEER, University of Bayreuth, 95440 Bayreuth, Germany
Introduction:
Scientific rationale:
Disentangling the various environmental factors driving emergent plant community properties such as productivity is a central goal in community ecology. Here, we propose a new phytometer experiment, which is intended to accompany on-going studies that manipulate environmental drivers of plant communities either through experimental manipulation or along natural gradients such as elevational ranges, countries and continents. We aim to develop the phytometer as a living reference system to be used as a tool for community ecology to improve ecological theory and prediction.
In order for a standardized phytometer to be a useful tool for community ecologists, several conditions should be met. First, a phytometer for community ecology must itself be a community of interacting species. Community effects can shape species-level outcomes in response to climate change, effects that would otherwise be missed by single species phytometers. Second, because of unique interactions between species, climate, and soil, a standardized phytometer gains efficiency when paired with a standardized substrate, which can then be compared to local soils to disentangle of climatic and edaphic conditions. Thus, an ideal phytometer is composed of a standardized plant mixture grown in standardized substrate. Our phytometer allows for the quantification of climatic and edaphic effects on plant growth independently from one another.
We introduce a phytometer consisting of a three-species mixture representing common European weeds, which are naturalized but non-invasive on six continents, combined with an inert standardized substrate mixed with a standard amount of fertilizer. Our goal is to provide researchers a
29
common response variable that is independent of local soil resource pools and regional species pools. By providing such a common metric, emergent community properties such as productivity can be gauged relative to other sites that have implemented the same study. The protocol is streamlined for relative ease, and undemanding in terms of effort required.
Guidelines for participation:
The phytometer as a tool is intended to accompany experimental or observation studies on herbaceous plant communities that operate across gradients with turnover in soils and species. This could mean distances between phytometer sites of 5 meters at an experimental site, or across countries and continents for large-scale gradient studies. We are also interested in single site phytometer installations to build a larger phytometer database in order to establish a geographic network of phytometer performance. Our phytometers are designed for herbaceous, grassland systems so should be placed in an area of full sunlight. The phytometer initiative is intended to accompany, not replace, ongoing studies that manipulate environmental drivers of plant communities either through experimental manipulation or along natural gradients.
We have designed the phytometers to be low cost and low effort. While we are able to provide the basic supplies for phytometers, there is some cost required by local sites in order to obtain additional environmental data. This is detailed in the ‘Materials required’ section, but we anticipate the cost being ~€550 per site plus shipping costs, with additional sites per group being
~€350. Labor requirements for each site consist of growing phytometer species from seeds, filling phytometer pots with soil and transplanting species into them, and two harvest dates separated by one year. Phytometers should be planted so that the first 50-days of growth coincide with peak growing conditions of the site, as determined by local investigator expertise. Note that there are 66 days of greenhouse growth required prior to this 50-day period. Participation requires that sites have access to local climate information, with daily resolution for minimum and maximum temperature, and daily precipitation. This document provides a short overview of the timeline of events, and a detailed section describing a step-by-step procedure.
30 Abbreviated protocol (follow line numbers for more details): 1.1) Receive package from Bayreuth with all materials needed for growing and planting the phytometers.
1.2) Order additional materials – loggers and PRS probes
2.1) Begin germinating seeds in potting soil in trays in greenhouse; keep soil moist at all times.
2.2) Transfer individuals to quickpots; 1 per cell.
2.3) After three weeks, move quickpots outside to harden off. Keep soil moist at all times.
3.1-3.2) Fill vermiculite phytometer pots and water.
3.3) Retrieve local soil from site and fill pots to the rim. Insert in-growth cores into all pots.
3.4) Dry teabags in 70°C oven for 48 hours. Weigh.
3.5) Refill vermiculite pots to rim. Mix in Osmocote fertilizer evenly across pot in the top 1-3cm.
3.6) Transplant seedlings to phytometer pots using the planting scheme, planting disc, and stamping tool.
3.7) Weigh non-planted individuals to determine average starting biomass.
3.8) Attach labels to phytometer pots.
3.9) Install add-ons: teabags, in-growth cores, and TidbiT dataloggers.
3.10) Begin 10 days of watering; 1L per pot per day
3.11-3.13) Move pots to field site, dig them in up to the rim in checkerboard fashion.
3.14) Insert PRS probes into local and standard soil. 4.1-4.2) Measure maximum vegetative height of all living individuals. Record mortality and number flowering/seeding individuals in each pot.
4.3) Harvest aboveground biomass in pots. Individuals should be clipped 3cm above ground. Dry biomass at 60C for 48h. 4.4) Remove root in-growth cores, extract roots from soil, replace cores with root free soil/substrate.
4.5) Remove PRS probe and ship.
4.6) Dig up dataloggers, download data, and rebury. 4.7) Weed out non-phytometer species
5.1) Weed phytometer of any non-phytometer species in spring of next year.
5.2) One year after first harvest, repeat steps from first harvest, except for PRS probes.
5.3) Remove teabags, dry in 70C oven for 48 hours and weigh.
Figure 1: Workflow for key steps. Dates are referenced to the first harvest event, which should closely coincide with peak growing seasons on a site-by-site basis
Begin 10 day watering period L per day per pot
1
74 days before harvest
10 days before harvest days before harvest
50
120 days before harvest
116 days before harvest
88 days before harvest
60days before harvest 62 days before harvest Fill standard substrate pots ;
Retrieve local soil ; insert in-growth cores
Insert PRS probes Prepare phytometer pots
Record height; mortality . Download logger data; rebury Remove PRS probes Order additional materials
Plant seeds for germination
Transplant seedlings to phytometer plants. Weigh non
-planted individuals. Label pots; install loggers
Move pots to field site; Start of no maintenance
Harvesting aboveground biomass 3cm above soil . Extract root ingrowth core
Weed after harvesting
Record height ; mortality. Harvest biomass, download logger data. Extract root in
-growth core. Place teabags in drying oven
for 48 hours and weigh Quickpots with individuals
Harvest date / local peak growing season
End of first year Weeding in spring Site dependent
31 1. Materials required:
In the following list of required materials, some line items are labeled as ‘per site’. A site may be each observational area along an ecological gradient, or each treatment within an already existing experiment.
1.1 Provided by Bayreuth team:
1.1.1) Seeds of three target species.
1.1.2) 3 trays for germination and 3 Quickpots for early growth (separate by species) 1.1.3) Greenhouse potting soil for germination and Quickpot growth
1.1.4) 10 black plastic pots (30 cm inner diameter, 23 cm height) per site 1.1.5) Planting scheme
1.1.6) Planting disc (Allows planting at uniform density and size).
1.1.7) Stamping tool (Ensures plant plugs are inserted at uniform depth). 1.1.8) OSMOCOTE Exact Standard 12-14M” slow-release fertilizer. Composition:
NPK(MgO):
15+9+11+2; Amount: 4 g / standardized substrate pot.
1.1.9) Rooibos and green teabags for decomposition rates, 10 of each per site. 1.1.10) Plastic mesh ingrowth cores for belowground biomass. 10 per site. 1.1.11) Plastic cylinder for installing in-growth cores.
1.1.12) 100 liter bag of Vermiculite per site (Vermiculite G: K1 - 0-2mm grain size). 1.2 Items that must be ordered by site investigator. Links are provided to the company,
which can link you to your nearest distributor:
1.2.1) TidbiT dataloggers. Two required per site. Ordered at:
o http://www.onsetcomp.com/products/data-loggers/utbi-001
1.2.2) TidbiT dataloggers require a device to download data and computer software. This is a one-time cost for your group (i.e. not per site). Note: The software is used for all HOBO logger products, so please check if your lab group already has this.
o http://www.onsetcomp.com/products/communications/base-u-4 o http://www.onsetcomp.com/products/software/bhw-pro-dld (also
available as CD)
1.2.3) Two pairs of PRS probes per site. PRS probes can be ordered here: o https://www.westernag.ca/innovations/customer/order 1.2.4) Paper bags (for harvested plant mass).
1.2.5) Access to 60˚C drying oven.
32
2. Planting and greenhouse management:
Figure 2a-b: Material for steps 2.1-2.3 2.1 Sow seeds of each species in separate trays (Fig. 2a) with the provided potting soil.
Each species should have its own tray, with about 10g seeds per tray (Approximately double the number of individuals needed to account for germination rates). This should be done 116 days prior to the planned date of harvesting the phytometers. After sowing, lightly press the seeds into the potting soil and cover them with a thin layer of soil to prevent desiccation. Trays and quickpots should be watered daily such that the soil never dries out. However, the soil should never be oversaturated for more than an hour (i.e. do not over water). During the germination period, the temperature of the greenhouse should be between 18°C and 20°C. If a greenhouse is unavailable, this step can be done in a normal room that is temperature maintained and has natural light, though this should only be done as a last resort.
2.2 After 4 weeks of germination, transplant the individuals into the quickpots (Fig 2b) with one individual per cell. Fill the quickpots with the potting soil sent to you, with a small hole lightly impressed at the center of each cell. Carefully remove
individuals from the germination tray and lightly press them into each quickpot cell; make sure that the roots are completely covered by soil at the end.
2.3 After three weeks of growth in the quickpots, move them outside of the
greenhouse (or into a cold or temperate greenhouse if there is the risk of frost) in order to harden off. After 1 week of hardening off, transplant the individuals into the phytometer pots.
33
Figure 3: Left to right: Trifolium pratense, Plantago lanceolata, and Dactylis glomerata in Quickpots prior to transferring to phytometers.
3. Pot installation:
3.1 This step must occur during steps 3.2 and 3.3. When filling pots, first fill them with a small layer of soil or vermiculite. Using the planting disc, place the provided plastic cylinder into one of the three holes surrounding the ingrowth cores location (place ‘C’; see Fig. 5 below). Using the cylinder as a reference, use the stamping tool to create a hole for the ingrowth cylinder and place it in the soil. Hold the ingrowth core in place while you fill soil around it, taking care to not get local soil in the core. Fill the open ingrowth core, a funnel is helpful here. For standard substrate pots, it should be filled with vermiculite. For local soil, it should be filled with root-free soil. This means you must sieve a small amount of local soil free of roots (approximately 1 liter per site). The top of the in-growth core should be even with the local soil when finished, but vermiculite should be filled to the rim, burying the ingrowth core for now, as it will shrink over time. A plastic tag or stick can be used to mark the location.
Figure 4b: Preparing vermiculite pots. The left most picture is a pot that was filled to the rim and allowed to settle for two weeks.
3.2 Two weeks prior to planting the phytometers (74 days before harvest), fill the five standard pots provided by the Bayreuth team with vermiculite substrate (with ingrowth cores; step 3.1) and soak them with water. The vermiculite naturally settles and is an unwanted effect; this step along with 3.5 helps minimize settling after planting (Fig. 4b). The pots have holes in the bottom to allow water drainage, but place a paper towel at the bottom of the pot to prevent Vermiculite from spilling out prior to planting.
34
3.3 Transport local soil from your site to the greenhouse (~80L per site; can be stored in greenhouse until planting). Ideally, soil is coming from 5 - 30 cm below the surface, which is thoroughly mixed prior to potting. This soil should not include any green plant material or leaf litter. Fill soil up to the level of the in-growth core (2-3cm below the rim). If it is not possible to transport soil to the greenhouse, pots can be prepared on-site but please make note of it.
However, this requires daily visits to the field for ten days post-planting to water in the plants.
3.4 Label the teabags (follow labeling from step 3.6). At least two days prior to planting (62 day before harvest), place teabags in a drying oven at 70°C for 48 hours. Remove, weigh the teabags to the nearest milligram and record this weight. Removal and weighing should be done shortly before planting.
3.5 Refill vermiculite to the rim in standard pots, compress with your hands, and refill it again Add 4g of Osmocote provided by the Bayreuth team to the top of the
Vermiculite, and gently mix in so that it is still within the top 1 - 3 cm of the pot. Water the vermiculite and local soil as this makes planting easier. (Fig. 4b).
Figure 4c: Transplanting species to phytometer pots.
3.6 At this step, take care that you do not plant non-target species (see Appendix 1) as seed mixes often contain a low rate of contamination from other species. Place the planting disc over the top of each pot, align the small notch in the disc with the hole drilled into the rim of the pot, and press holes into the soil using the “stamp”. The notch aligns with the ‘top’ of the planting scheme. Place all holes prior to planting to keep the distance between individuals consistent. Once 18 holes are created, remove the disk, carefully remove healthy individuals from plugs and press them into the holes, making sure that no roots are exposed above the surface. While only healthy individuals should be used, avoid taking the biggest individuals first. Rather select in a ‘typewriter’ fashion (left to right, top to bottom), skipping over nonhealthy individuals (browning, considerably smaller, etc). Cover the potting soil from the plug with the soil / substrate from the pot and gently press around the stem of the plant to minimize desiccation (Fig.4c).
3.7 Measure the natural maximum vegetative height of 25 random, but healthy, individuals of each species that remain in the quickpots (i.e. not planted). Then, clip
35
these individuals at 3cm above the soil, place them in paper bags in groups of five, and dry them at 60˚C for at least 48 hours, and weigh them to the nearest
milligram. This provides an average starting mass for species by site.
3.8 Affix the provided labels to the small hole drilled in the pot rims with the following labeling scheme (Fig.4c): Site code - Soil type code - pot number. Site codes are two digit codes assigned to you by the Bayreuth group; soil type is either LS for local soil or SS for standard substrate, and pot number is 01-05. As an example, the third replicate of local soil in Bayreuth would be labeled ‘BT-LS-03’. These labels must be identical to those entered when reporting data. Pots labeled -03 should be those with the data loggers.
3.9 Dig a small hole for both teabags in all ten pots at 5cm depth for measuring decomposition rates, insert a teabag, and cover. Data loggers should be placed at 5cm depth in one pot each for local and standard soil (see Fig 5 for placement of all items). Make sure to activate loggers before planting at a 30 minute logging interval! Tie a piece of sturdy string to the data loggers and affix them to the hole drilled for the label in the rim of the pot. Care must be taken at this stage not to injure the plants.
Figure 4: Planting scheme and pot location for add-ons. Planting disc creates holes only for plants
3.10 After planting, water in pots for 10 days at a rate of 1 liter per pot per day.
36
3.11 Transport pots to the site, taking care that they stay upright. Settling of the soil and substrate may occur during transportation. Pots with local and standardized soil should be placed in a checkerboard fashion, 5x2, as space allows [Fig.5]. Align the pots so the hole on the rim with the label faces north. Once all pots are placed in the ground, fill in the soil surrounding the pots and compress to ensure insulation. Extreme care should be taken to prevent local soil from getting into the
standardized Vermiculite substrate phytometers at this step. We suggest inverting an empty pot over the standardized Vermiculite substrate pots to accomplish this. Please photograph each site with the arrangement of all ten pots visible, and send to the Bayreuth Group.
3.12 If alternate pot arrangements are required to fit the available space, this should be diagrammed and sent to the Bayreuth Group. In the event that separate blocks are required
due to space constraints, standardized substrate and local soil phytometers should be present in each block.
3.13 This is the start of the 50-day growth period for the phytometers. No maintenance of the phytometers occurs at this stage.
3.14 PRS probes should be inserted 10 days prior to harvest . Samples consist of four pairs (anion/cation probes). One pair should be inserted into phytometer pots 01-04 for each soil type (points B and C; Figure 4). Insert probes vertically, with the flat side facing the rim of the pot by making a slot in the soil with a small spade or soil knife and pushing in the probe pointed-side down, leaving the top 2cm of the probe exposed. It is important to cut deeply enough into the soil, as the PRS probes will break if you try to force them through the soil. Once buried, make sure there is proper contact between the probe and substrate by firmly pushing the substrate back around where the probe was inserted. No substrate should be removed during this process.
37
Figure 5: Preferred arrangement of pots in checkerboard fashion (left).
Alternate arrangement showing one block necessary to fit in footprint of experiment (right)
4. Harvesting
4.1 Prior to harvesting, record the highest, naturally standing vegetative height of each individual to the nearest millimeter.
4.2 Record number of individuals of each species in each pot that are A) still alive and B) have reproductive biomass.
4.3 To harvest biomass, pull each individual erect and then clip 3 cm above the soil. Place the living biomass of the same species within each pot in one paper bag, and senesesced or dead biomass of the same species in a separate paper bag. Thus, each pot should have six bags, and each site should have 60 bags. Reproductive biomass should be included with living biomass. Dry bags in a drying oven at 60˚C for at least 48 hours, and then weigh to the nearest milligram, or most accurate mass available. Once weighed, retain biomass in case of future leaf chemistry analysis. Enter height, survivorship, and biomass data into the provided data sheet and send to Peter Wilfahrt (peter.wilfahrt@uni-bayreuth.de).
4.4 Carefully remove in-growth cores from the soil without uprooting any plants by cutting around the mesh with a sharp knife. Wash the roots free of soil over a fine sieve (see Appendix 2). Collect all roots per pot in a paper bag, label it, dry it at 60˚C for at least 48 hours, and weigh it to the nearest milligram. The root free soil is replaced in the pit so a subsequent harvest can take place after one year. For vermiculite, this can be replaced from the bag. For local soil, make sure the soil is root-free by sieving over 2mm mesh prior to replacing.
