@phdthesis{Smiljanic2019,
  author    = {Marko Smiljanic},
  title     = {Growth responses of peatland pines : can peatland Scots pines (Pinus sylvestris) be used as dendroecological proxies to peatland hydrology and environment?},
  journal    = {Wachstumsreaktionen von Torfkiefern : k{\"o}nnen Torfkiefern (Pinus sylvestris) als dendro{\"o}kologische Stellvertreter f{\"u}r die Hydrologie und Umwelt von Torfmooren verwendet werden?},
  url       = {https://nbn-resolving.org/urn:nbn:de:gbv:9-opus-33252},
  pages     = {74},
  year      = {2019},
  abstract  = {Peatlands are wetland ecosystems covering a relatively small area of the World (~3\%), but at the same time storing excessive amounts of carbon for a very long time (equivalent to the four times global annual net primary production). As carbon sinks, peatlands work in spite of their slow growth, absorbing carbon dioxide (CO2) through the photosynthetic activity of the peatland plants and their low growth rates, and because high groundwater table removes oxygen from the soil and slows down the decomposition of the dead plant matter. Because of the relative lack of the oxygen in the peat, especially compared to the mineral soils, methanogen populations in the peatlands are abundant, and releasing methane (CH4), a potent greenhouse gas, to the atmosphere. Therefore, peatlands are generally at the same time significant carbon sinks and stores as well as the methane sources. The balance among the two peatland gass fluxes (CO2 and CH4) will dictate the impact of any given peatland on the global climate and primarily driven by hydrology, in the form of the groundwater table levels. Because of the slow decomposition rates, and from radiocarbon dating of the peat as well as the subfossil records buried in it, carbon stored in peatlands is locked for a very long time (centuries to millennia). It is, therefore, crucial to gain insights into the development of peatlands and their gas balance through time. One way to get both is by studying peatland hydrology in the form of the groundwater table levels and their historical variations. Unfortunately, intensive monitoring of peatland groundwater table, when available, is an only a recent endeavor. Therefore, we need to employ proxies to reconstruct the past by leveraging the present. In statistics, proxy variables are often used when the observations of the variable of interest, are either missing or too difficult to obtain. In this thesis, I tested whether we can use the radial growth of the Scots pines growing on peat as proxies to the peatland hydrology. To that end, I studied growth responses of the peatland Scots pines. Other proxies can and are used for the reconstructions of the groundwater table levels, but tree-growth is widely used as one of the proxies to reconstruct past environments which is at the same time annually resolved. First, I examined the growth ecology of the peatland Scots pines by looking at their intra-annual development and trying to find relationships between it and environmental factors while at the same time comparing it with the Scots pines growing at the forest sites. I first tried with wood anatomy and found that, unfortunately, peatland Scots pines do not form enough wood cells, and consequently do not have high temporal resolution, necessary to investigate the intra-annual patterns of the radial growth. Initial results from wood anatomical investigations were interesting none-the-less, indicating that peatland Scots pines might have smaller cell features than the Scots pines from forests, but might at the same time maintain Early/Latewood ratios of those same features. After I found that wood anatomical series were not resolved enough I decided to go with dendrometers, linear displacement sensors which constantly monitor the variations of stem radius, to get insights into the intra-annual growth patterns of the peatland Scots pines. Before using dendrometers for ecological investigations, I was involved in implementing routines commonly used in the analysis of the dendrometer signals and bringing them to R in the form of the dendrometeR package. At one peatland complex, I installed dendrometers on ten trees in total at both peatland and forest sites and compared the pattern of the standardized signal. I inferred from the comparisons and classifications that the signal from two sites was indistinguishable for the dendrometer series shorter than five days. Furthermore, the most important environmental factor driving the radial variation at the peatland site was hydrological, daily relative humidity, indicating further that peatland hydrology might indeed be the driver behind peatland Scots pine growth. Finally, I looked at the growth responses of peatland Scots pines from central Estonia using dendrochronological methods. Peatland hydrology, in the form of the groundwater table levels, was indeed the environmental factor with the strongest, and also stationary, correlations with the radial growth of the peatland Scots pine. That relationship indicated that peatland Scots pines are indeed possible proxies for reconstructing past levels of the peatland groundwater tables. My study further indicated that the growth response of the peatland Scots pines was non-linear, further complicating the reconstructions of the past peatland hydrology. However, the strength of the growth response was proportional to the general hydrological regime, expressed as median groundwater table level. As the hydrological regime of the peatland does not vary considerably on the annual scales, but more on decadal it might be more appropriate to find another, independent, proxy to the hydrological regime first, and than use annually resolved radial growth of the peatland Scots pine to reconstruct past levels of the peatland groundwater table.},
  language  = {en}
}