Cycling of Carbon and Other Elements in a Beech Forest Hestehave, Jutland, Denmark, in the Past 50 Years

Plant biomass, primary production and mineral cycling in the beech forest (Fagus sylvatica L.), Hestehave in Jutland, Denmark were studied over a 50-year period. The role of the forest as a carbon sink was also assessed. Aboveground tree biomass was 226 t∙ha−1 in 1970 and after a 50year 539 t∙ha−1 in 2014, an unexpected increase with 313 t∙ha−1. Annual production at those two points in time was 13.4 and 20.5 t∙ha−1, respectively. It was apparent that the tree biomass was still acting as a sink for carbon, which was the dominant element in the aboveground parts. The concentration of other elements (N > K > Mg > P > S > Na > Mn > Zn > Fe > Cu) ranged from 495 to 0.4 kg∙ha−1. Annual litterfall restored 3.2 t∙ha−1 to the soil as organic matter or 1.6 t∙ha−1 as carbon. Over the year 53% of the litterfall was decomposed. A pH decrease of 0.95 units in the soil was observed between 1968 and 1993. This was attributed to fallout from a neighbouring thermal heating station affecting sulfur deposition and increasing soil acidification. After 1993, when filters were fitted in the heating station, the pH decrease in the soil was smaller, only 0.09 pH-units up to 2011. The increased tree growth is an additional, likely explanation for the observed soil acidification. Deposition of the growth-limiting element nitrogen increased during later years and is now, most likely around 20 kg∙ha−1 per annum, which may partly contribute to the increased production.


Introduction
The effects of climate change on forests may result in a major environmental issue, e.g.changes in processes such as forest photosynthesis, decomposition and mineralisation may affect the turnover of carbon and minerals?The question is whether the forest will become a sink for carbon mitigating climate change or a source.The carbon balance of forests is therefore an area of major concern.Data are needed to draw up budgets for carbon and other elements in the ecosystem.
This study analysed data obtained in investigations in a beech forest (Fagus sylvatica L.) in Hestehave, Jutland, Denmark, established as a part of the Danish International Biological (IBP) programme in 1967.The focus of these investigations was on primary and secondary production in forest including decomposition (See Annex 1 for major publications).
The values for tree biomass, tree productivity and mineral cycling of elements reported here have earlier only been available in an international data bank (Andersson, 1973;Thamdrup, 1973;Reichle, 1981).The mineral cycling could be addressed as other scientists made their data available to the author.Another major reason for this late publishing is that the problems actual today make the data useful and valuable as references especially in combination with the repeated measurements.New light can be shed on today's problem.Repeated measurements of tree diameter and height were therefore made in November 2014.This made it possible to discuss the changes in tree biomass and production taken place during a 50-year period.
The paper reports preliminary findings on tree biomass, production and cycling of elements and discusses possible changes in these properties in an almost 50-year perspective of these.Special emphasis is placed on carbon, nitrogen and sulfur.

Research History
Danish beech forests represent a classical subject in the research history of production ecology.In today's textbooks the classical diagram of Möller et al. (1954) can be still found indicating the different components of gross and net production over time in a beech forest.Furthermore the father of production ecology was a Dane-Boysen Jensen (1932).This investigation of a Danish beech forest has therefore a strong historical background.

Hestehave-A Beech Forest
The study site, comprising 3 ha, is situated in a pure beech forest (Fagus sylvatica L.).At the time of establishment of the research programme in 1967 the forest was 85 -95 years old with an overstory tree density of 210 trees•ha −1 , average tree height of 27.9 m and diameter at breast height 40.5 cm (Table 1, Figure 1).There was also an understory with 186 trees•ha −1 and an average tree height of 10.6 m.A review of part of the study site was done in November 2014 revealed some wind damage in the eastern part of the area.Some old trees, living or dead, had fallen, but the damage was restricted.Three areas (D-F/7-9, A-E/14-16 and A-B/11-12 according the original map) were identified and remeasured for tree height and diameter.They had a combined area of 0.28 ha and a tree density of 189 trees•ha −1 .The most striking changes over the almost 50 years were a diameter increase of 17.5 cm and a tree height of 6.2 m.Compared with the early observations, visible changes to the forest had occurred.For example, dead material had accumulated on the ground and a previously described and investtigated understory layer of beech had decreased with most of these trees having died or merged into the over story tree layer.However, the experimental area as a whole was less altered than expected and could be used for the reinvestigation.
The soil type at the site is dominated by an Ortic Luvisol.It is characterised by an underlying calcareous or agric B-horizon with clay accumulated after leaching or downwash (Dalsgaard, 1983).Another common name  could be "Parabraunerde".The pH-range for the upper 10 cm of the soil at the start of the project in 1967 was 4.2 -7.6 and base saturation was 13% -89%.
In spring a dominating species and aspect in the Hestehave forest is Anemone nemorosa, which is later followed by Asperula odorata, Carex sylvatica, Circea lutetiana, Hordeum europium, Melica uniflora and Veronica montana.

Element Cycling Model
In a previous paper, an element cycling model adapted from Nihlgård (1970Nihlgård ( , 1972) ) was used (Andersson, 2014).For the purposes of synthesis and comparability, the same model was used here without further adaptations.

Methods
The methods applied in this paper refer to those, which have been used by the author previously.References to earlier publications are given.Some data have been put to the author's disposal.Where essential differences exist methodologically references to additional papers are given.In short, key areas and publications are as follow:

Biomass and Production of Trees, Shrubs and Field Layers
The main methods used for describing trees and shrubs can be found in Andersson (1970c;2014) and Nihlgård (1970Nihlgård ( , 1972)).For trees and shrubs, non-destructive measurements as diameter and height were correlated in allometric regressions with destructive measurements on above-and belowground fractions.In order to make future measurements and calculations possible, the allometric regressions are given in Table 2. Eleven beech trees were sampled and measured for stem wood, bark, branches and currents twig biomass.Three sizes of stumps were excavated.Roots > 2 cm diameter, roots 0.5 -2 cm and <0.5 cm were also excavated in 10 pits 50 × 50 × 50 cm.Age and tree production were estimated by yearring analysis for 5-year-periods on stem discs.For the field layer see Hughes (1975), Bülow-Olsen (1977) and Astrup & Bülow-Olsen (1979).In the reinvestigations 2014 a tree measurement tape was used for diameter at breast height while for tree height a height meter SILVA Type 65 was used.
Table 2. Log 10 regression data on dry weight at (85˚C), where D is tree diameter in cm 1.3 m above ground and H is tree height in m of the Danish IBP-PT beech forest Hestehave, Kalö, Jutland, Denmark.These regressions were used when calculating tree biomass and production for 1967.

Elements and Their Distribution in Biomass and Production of the Shrub and Field Layers
The main methods used for determining element concentrations and then their distribution in plant material are described in Andersson (1970aAndersson ( , 1970b;;2014) and for trees and shrubs and in Bülow-Olsen (1977) for the field layer.For chemical values in plant material for 2014 data from 1970 was used assuming the same composition today and applying an adjustment for the weight increase.

Elements of Litterfall and Surface Litter
The litterfall of leaves, bud scales, flowers, twigs and miscellaneous items and the degradation of leaves as well as consumption of leaves were determined in 48 litter traps with a diameter of 60 cm placed at random and collected monthly (Nielsen, 1977).Larger twigs, branches surface litter amounts were measured in six 10 × 10 m squares.Chemical analyses were carried out according to Bülow-Olsen (1977).

Elements in Incoming Precipitation, Throughfall and Interception
The incoming precipitation was measured in two rain gauges outside the forest.The throughfall was measured in 24 rain gauges with an opening diameter 19.0 cm placed randomly within the observation plot.Stem flow was also measured on five representative trees.Chemical analyses of water samples followed Bülow-Olsen (1977).Important information on deposition rate of sulfur and nitrogen was lacking and were therefore derived from Jensen (1962), Jörgensen (1978) and Pedersen et al. (2001).Comparisons with south Swedish conditions were made.

Soil Organic Matter Exchangeable Mineral Content and Total Mineral Content of Trees
Soil sampling and chemical analyses followed Bülow-Olsen (1977).Analyses of tree material follow Andersson (1970c;2014).

Statistics
Results are generally presented as means.Percentage errors of the mean is given at 95% confidence limits.Andersson (1970b) discusses different statistical aspects of errors in sampling and computations.

Principles for Investigations of Changes in Tree Biomass and Productivity over Time
A reinvestigation of a forest stand for tree biomass and production would require that the very same stand or sample plots and trees are revised again.In this reinvestigation three small sample areas of the original experimental area were measured representing 0.28 ha, compared to three areas of 1 ha in the original study from 1967.
A map of the experimental area was available and old tree numbers and location could be identified.Unfortunately, the original data of tree diameter and height were not available.Therefore the criteria for a revision was not completely fulfilled.However, the original allometric regressions for tree biomass and production were used for new calculations using the diameter and height data from 2014.A reservation for the shortcomings at the interpretation is however expressed as error could have been introduced at the comparison.

Results
6.1.Biomass and Production of the Tree, Shrub and Field Layers

Initial Results 1970
The organic matter in terms of biomass and production by the tree, understory, shrub and field layers are essential carriers of mineral elements, in particular carbon.The aboveground biomass of the beech overstory in 1970 was estimated to be 220.3t•ha −1 (Table 3).There was also an understory, often with small trees of beech, corresponding to 5.8 t•ha −1 .As regards the belowground system the stumps and associated roots of the overstory had an estimated tree biomass of 38.9 t•ha −1 .An additional 3.8 t•ha −1 of stumps and roots was estimated for the understory giving a total of 42.7 t•ha −1 for the belowground system of beech trees.This gave a total of 268.2 t•ha −1 of tree biomass for the beech forest.
In terms of annual biomass productivity a rate of 13.65 t•ha −1 was found for the aboveground parts of the overstory.Examination of the distribution into different fractions revealed that the understory of beech yielded an additional 1.13 t•ha −1 annually.The production of the belowground parts was at a low estimate of 2.37 t•ha −1 .However, this is known to be an underestimate and needs to be interpreted as such, as the methodology used only considered larger roots and the rapid turnover of finer roots or hairs are were overlooked.The total productivity of the beech forest was 17.15 t•ha −1 .
The total plant biomass and the annual production rate of the field layer was marginal relative to those of the tree layer (Table 4).The productivity was estimated to 1 066 kg•ha −1 , with 55% and 45% aboveground and belowground, respectively.
The tree biomass was also determined as carbon (Table 5).A total of 103.7 t•ha −1 was found for the aboveground parts including both overstory and understory.An additional 18.8 t•ha −1 was provided by the belowground parts giving a total of 122.5 t•ha −1 .The field layer was also considered.

Results 50 Years Later
Based on results from other investigations in a mature mixed oak wood in southern Sweden (Andersson, 2014) it was concluded that Hestehave beech forest was most likely still in an aggrading phase accumulating biomass in trees.This indicates that the forest ecosystem was acting as a sink for carbon.This fact became a hypothesis worthwhile testing on other forests.The ongoing publishing of the Hestehave work was then in November 2014 combined with a revision of tree diameter and height in 2014 in order to test the hypothesis.
The most striking feature of the beech forest when revisited after almost 50 years was the increased tree diameter and height, changes which were visible for the naked eye (See Table 1, Figure 2).The tree biomass had increased from 220 to 539 t•ha −1 or by 154% and the tree production from 13.4 to 20.5 t•ha −1 or 53% (Table 6).Due to high costs and the uncertainties, only the aboveground parts were considered.In terms of carbon there had been an increase from 101 to 245 t•ha −1 or 142%.
Thus it was confirmed that the beech forest still is acting as a carbon sink when considering the aboveground parts alone.

Elements and Distribution in Tree Biomass, Production and Field Layer
In the aboveground parts calcium (Ca) and nitrogen (N) were the dominating elements with 867 and 495 kg•ha −1 , respectively, followed by potassium (K) > magnesium (Mg) > phosphorous (P) > sulphur (S) > sodium (Na) > manganese (Mg) > zinc (Zn) > iron (Fe) and cupper (Cu) ranging from 292 -0.4 kg•ha −1 (Table 5).In the belowground parts the order was N and N followed by the elements K > Mg > P > Na ranging from 113 -18 kg•ha −1 and minor amounts of the other elements analysed.

Elements of Litterfall and Surface Litter
In functional analyses the forest litterfall and surface litter are essential components (Table 7).The litterfall in 1970 had an organic matter content of 3230 kg•ha −1 corresponding to 1620 carbon kg•ha −1 .This amount was returned to the soil in that yearly.However, there was a great variation between years as in 1971 the corresponding amount was estimated to be 4030 kg•ha −1 .The other elements were returned to the litter layer in the following order N > Ca > K > Mg > Na > P ranging from 34 -1.7 kg•ha −1 .The surface litter consisted of branches and smaller fraction comprising 7490 kg•ha −1 from which more Ca was returned more than N (Table 7).

Elements in Incoming Precipitation, Stem Flow, Throughfall and Interception
Input of elements by rain to forest is important.Here it was measured in an open field (In).Some of the rain and its elements fall through the tree canopy as throughfall (T) and some follows the trunk of the trees as stem flow (ST).There is a difference between what comes in and what reaches the soil (Diff) and this difference can be negative or positive.For the precipitation a negative difference means that water has been lost by interception.For elements a negative value is a result of uptake, while a positive value means a leaching of elements from the canopy.Deposition of elements in forest occurs in two forms, wet and dry, where the later refers to aerosols in some deposited in one way or the other.The yearly deposition in the Hestehave beech forest was followed for one year.Data for Na, K, Ca, Mg, Mn and P are given in Table 8.Two important elements were missing, S and N.In the period between 1970 and 2014 major changes in air pollution have taken place in the Nordic countries, especially for S and N components.Although data on these changes of these elements were lacking for Hestehave  forest it was deemed important to describe changes as they are important for a long-term maintenance of biomass production.Some old and new information from the area was therefore synthesized (Table 9).
The nearest measurement station to Hestehave is Ödum.There 12 kg•ha −1 S and 7 kg•ha −1 N was measured yearly around 1960 in an open field (Jensen, 1962).For the open field in 1970 the amount of S could be estimated to be higher and N only slightly higher.The throughfall in the Hestehave forest then could have had an S content of 15 -25 kg•ha −1 yearly.We assume that the tree growth was limited by nitrogen and therefore the throughfall contained less N than the incident rain as an uptake in the canopy took place.This is supported by data from southernmost Sweden and general knowledge.Information from Skovbjerg indicates realistic values of today.Values from Sepstrup Sande/Tange give the present situation of deposition for Hestehave with still less S and more N.A decrease in soil pH (measured in water solution) was found for Hestehave between 1968 and 1993 of 0.95 units (Dalsgaard K in Pedersen, 2011).After 1993 the decrease was smaller, only 0.09 pH-units up to 2011.It was concluded that there was a thermal heat plant in the vicinity of Hestehave and deposition of acidifying substances impacted the Hestehave forest.After 1993 the impact was smaller as filters were introduced at the plant.

Soil Organic Matter and Exchangeable Mineral Content
The organic matter content of the soil (SOM) down to 80 cm depth was estimated to be 141 t•ha −1 or 71 t•ha −1 as carbon and the nitrogen content was 7.5 t•ha −1 (Table 10).The exchangeable (Exch) amounts in an extraction with ammonium acetate (Am-Ac) were in the order decreasing order: Mn > Ca > Fe > N > K.It would be possible to estimate soil respiration, using the annual temperature of the soil to predict the rate of soil respiration (Bahn et al., 2010).
The dynamics of the elements in the system is seen in Table 11.The elements have different turnover times.Sodium and nitrogen are turned over more rapidly than potassium, calcium, manganese and phosphorous.

Discussion
This paper focuses on total plant biomass, annual tree production and mineral cycling in the Hestehave beech forest in Jutland, Denmark.Special attention was paid to beech tree biomass and production as almost half a century of changes could be assessed.Between 1970 and 2014, total aboveground tree biomass increased from 226 t•ha −1 to 539 t•ha −1 or by 313 t•ha −1 .In terms of tree productivity, annual biomass production rate, increased from 13.4 to 20.5 t•ha −1 in the same period.The yearly loss of biomass as litterfall was estimated to be around 6 t•ha −1 .The increase in tree diameter and height observed in 2014 was in line with this increase.
For other beech forests representing different fertility levels the total aboveground biomass has been reported to range from 225 to 315 t•ha −1 and the net annual production from 10.6 to 17.7 t•ha −1 (Nihlgård & Lindgren, 1977;Ellenberg et al., 1986).The values for the Hestehave forest in 1970 fall within this range.However, the tree biomass and productivity in 2014 were unexpectedly high.During the last 50 years the beech forest has been in an aggrading phase accumulating carbon in the aboveground parts at a yearly rate of 6.0 t ha -1 carbon.Around 57% of the litterfall was decomposed during a year.The available data did not allow any interpretations regarding the belowground system in the forest as a sink/source function.
The observed soil acidification could be explained by the deposition of sulfur compounds.A contributing factor could also be the extreme tree production during the last 50 years, which also implies that cations are taken up by the trees, which leads to an acidification (Ågren & Andersson, 2012;Tamm & Hallbäcken, 1988).References: 1. Andersson, 2014. 2. Nihlgård, 1972. 3. Jensen, 1962. 4. Pedersen et al., 2001. 5 Oral comm.T Ellerman 6.This paper.?Not measured.The observed increase in diameter, height plant biomass and production of the investigated beech forest in Hestehave was interpreted as a result of a changing environment over time.In a Danish study (Skovgaard & Henriksen, 1996) reported a significantly increased productivity expressed as increased tree height growth.App. a 3.6 m height increase was found during the period 1920s -1990.In an investigation led by European Forest Institute (EFI, 1996;1999) focusing on the importance of a changing environment it was found that an increasing growth was found in large parts of Europe.The reasons behind this could be different factors.Most likely the increased CO 2 level, increased nitrogen deposition and temperature were emphasized.The increased temperature implies also a longer growth period.
In a recent study of a number of European long-term forest experiments have been analysed (Pretzsch et al., 2014).It is concluded that the dynamics of forest growth has accelerated in Europe since 1870.European beech is one of the tree species treated and 22 experiments had been considered.36 Norway spruce experiments were also analysed.It was found that there is a faster tree growth (+32 to 77%), increased stand volume growth (+10 to 20%), increased tree biomass accumulation (+6 to 7%); tree numbers are lower and self-thinning remains constant.

Conclusion
In this paper the original results from 1970 on tree biomass, production and mineral cycling are reported.Of particular interest has been to elucidate if the Hestehave forest today is a source or sink for carbon.It was concluded that it is still a sink.This was revealed trough a new analyse in 2014 of tree diameter and height.Unexpected high values for increased diameter, height, tree biomass and production were found.In comparison with other data, the results seem most likely to be correct although a reservation has been given on methodological matters.A number of properties, which are changing in the forest dynamics today can be verified from the reported findings in the repeated Hestehave investigation.Financial resources were lacking for a more detailed analysis.New knowledge is needed on today's forests for their management.The value of long-term experiments needs to be stressed.

Figure 1 .
Figure 1."In the Hall of Freja" according to Adam Oehlenschläger-the author of the National Anthem of Denmark.The Danish Beech Forest at Hestehave, Jutland Denmark in 1967.Photo: H Petersen.

Figure 2 .
Figure 2. " In the Hall of Freja" according to Adam Oehlenschläger-the author of the National Anthem of Denmark.The Danish Beech Forest at Hestehave, Jutland, Denmark 47 years later in 2014.Photo: H Petersen.Table6.Tree biomass and production and carbon and nitrogen dynamics over a 44-year perspective in a European beech forest (Hestehave.Jutland, Denmark).Data for 1970 compared with data from 2014 based on a remeasured tree diameter applied in previous allometric regressions.Element content in 2014 was assumed to be the same as in 1970.

Table 1 .
Average values of stand properties of the Danish IBP-PT beech (Fagus sylvatica) forest Hestehave, Jutland, Denmark, in spring 1970.Diameter and basal area refer to breast height 1.3 m above ground.Measurement I: Data supplied by G. Münter, The Royal Agriculture and Veterinary University, Copenhagen.Measurement II: Average tree diameter and height in a selected part (0.28 ha) of the original investigation area in November 2014.

Table 3 .
Plant biomass and yearly production of the tree layer in the Danish IBP-PT beech forest Hestehave, Kalö, Jutland, Denmark.Figures given as metric t•ha −1 dry weight (85˚C).Standard error at 95% confidence limits.Production figures for 1967 represent the mean for the five year period 1966-1970.Biomass figures correspond to the growing period 1967.

Table 4 .
Yearly productivity of the field layer in kg•ha −1 at dry weight (65˚C) in the Danish IBP-PT beech forest in Hestehave, Kalö, Jutland, Denmark.Data for 1970.Figures supplied by A. Bülow-Olsen.A = Aboveground; B = Belowground.

Table 5 .
Weights of biomass and mineral contents in the tree and field layers of the Danish IBP-PT beech forest Hestehave.Kalö.Jutland.Denmark.Material sampled October 1970.

Table 7 .
Content of different elements in litterfall 1970 (L) and surface litter (SL) in the Danish IBP-PT beech forest in Hestehave, Kalö, Jutland, Denmark.Figures supplied by B. Overgaard Nielsen and chemical data on litterfall by M. Astrup.

Table 8 .
Precipitation and concentrations of elements in precipitation (In), throughfall (T), stem flow (SF) and interception (Diff.)for the period May 1970-April 1971 in the Danish IBP-PT beech forest in Hestehave, Kalö, Jutland, Denmark.Values supplied by M. Astrup.

Table 9 .
Sulfur and nitrogen deposition to open field and forests-a comparison of data from Jutland.Denmark and Southern Sweden showing changes over the last 50 years.Results given as kg•ha −1 yr −1 .Italics indicates best guess values.

Table 10 .
Distribution of elements in soil organic matter (SOM) and in fractions exchangeable in ammonium acetate (Am-Ac) and EDTA in the Danish IBP-PT beech forest in Hestehave, Kalö, Jutland, Denmark.

Table 11 .
Distribution of elements in some important aboveground functional fractions and turnover characteristics in the Danish IBP-PT beech forest in Hestehave, Kalö, Jutland, Denmark.