Plant Biomass , Primary Production and Mineral Cycling of a Mixed Oak Forest in Linnebjer , Sweden

Plant biomass, primary production and mineral cycling were studied in a mixed deciduous forest (Quercus robur L., Tilia cordata L. and Corylus avellana L.) in southern Sweden. Plant biomass amount above and below ground was 201 and 37 t∙ha−1, respectively. Primary production above and below ground was an estimated 13.3 and 2.3 t∙ha−1, respectively. Carbon was the dominant element in the forest ecosystem, comprising 133 t∙ha−1. Other major elements were: N > Ca > K > Si > Mg > S > Mn > P > Fe and Na (range 1123 to 18 kg∙ha−1), followed by some trace elements. Yearly litterfall restored 6.0 t∙ha−1 organic matter or 2.3 t∙ha−1 carbon. Approximately 45% decomposed and returned to the soil during the year. Monitoring of other elements revealed that the ecosystem received inputs through dry and wet deposition, in particular 34.4 kg∙ha−1 S and 9.4 kg∙ha−1 of N yearly as throughfall. Determination of yearly biomass increase showed that the oak forest ecosystem was still in an aggradation or accumulation phase.


Introduction
Over the years, the Linnebjer nature reserve, which lies 7 km NE of the city of Lund in southern Sweden, has been the subject of detailed investigations of its history, vegetation, organic matter, plant biomass, primary production, water profile, soil physics and soil chemistry.The studies on biomass and primary production led to the development of methods which were later used in Sweden and the Nordic countries.The methods used to determine plant biomass and production and the related results are described in Andersson (1970c) and Reichle

The Mixed Forest
The dominant vegetation in the Linnebjer nature reserve is mixed deciduous forest.Part of the reserve was used as a special area for detailed investigations (Figure 1 and Figure 2).A full description of the vegetation during the study period is given in Andersson (1970b).In brief, the forest has a number of layers: tree layer > 15 m, upper shrub layer 2 -15 m, lower shrub layer < 2 m and field and bottom layers.The tree layer is dominated by Quercus robur and Tilia cordata and the upper shrub layer by Corylus avellana.The field layer shows seasonal changes, with Anemone nemorosa dominating in spring and Oxalis acetosella in summer.The bottom layer is weakly developed.

Element Cycling Model
Cycling of elements was assessed using a model developed by Nihlgård (1972) (Figure 3).The analysis was limited to the above-ground parts, as root production and the release of nutrients from the root litter and leaching of elements from the soil were not estimated.However, root biomass was measured.In determining the dynamics of organic matter in a forest, the above-ground yearly net primary production (P) the yearly litterfall (L) and the decomposition rate (k) are the most important factors.Understanding element cycling requires information on the amounts of elements in the primary plant production, the biomass increase (ΔB), yearly litterfall (L), surface litter (SL) and the canopy-leached fraction (CL = the difference between total through fall and incoming rainfall).

Tree Biomass and Production
A detailed description of the methods developed and applied to the tree layer and the results obtained are provided in Andersson (1970c) and Nihlgård (1972).Therefore, only a summary is given here.The estimates of tree biomass and production were based on allometric regressions, using the relationship between destructive and non-destructive measurements.The estimation procedure comprised four basic steps: • Stand analyses of non-destructive measurements-tree height and diameter.
• Destructive measurements of sample trees: stem sections or discs, branches, twigs and roots-diameter, age.
• Application of the stand data to the regressions of the sample trees obtained in (2).
The above-ground biomass and production of trees were studied specifically for 11 Quercus robur, 6 Tilia and Sorbus and 18 Corylus.Studies of the below-ground parts with roots were limited to three individuals of each tree species.The main root biomass fraction, comprising roots > 0.5 cm in diameter, was collected around the tree.The fine root biomass (<0.5 cm diameter) was collected in 10 pits measuring 50 cm × 50 cm × 60 cm excavated along a transect at regular intervals of 2 m.

Field Layer
The above-ground biomass of the field layer was sampled on occasions corresponding to the maximum development of the seasonal species.Total primary production was calculated by adding the maximum values obtained for the different species investigated.These were sampled in 16 squares (50 cm × 50 cm) selected at random on each sampling occasion.Below-ground parts were collected within the same squares.

Litter and Litterfall
Litterfall was collected in 16 litter traps (50 cm × 50 cm × 30 cm) laid out in a systematic way.The trap sides were made of wooden boards and the base of nylon netting and they were placed 20 cm above the ground surface.Coarse litter such as twigs and pieces longer than 0.5 m were collected yearly in a 20 cm × 20 cm square sampling area.Random samples of surface litter were taken twice a year, before and after leaf shedding, in 16 squares measuring 50 cm × 50 cm.

Deposition
In connection with an investigation of element cycling in the oak forest (Andersson 1970b), rain samples were collected and subjected to chemical analysis in order to determine the deposition of sulfur and nitrogen, as well as other incoming nutrients.These samples were collected as through fall in a special area of the forest (Figure 1 and Figure 2).

Collectors and Collection
The incoming rainfall was measured at a nearby farm 300 m from the throughfall sampling area.A rain gauge of type SMHI with a diameter of 16 cm and fitted with a wind shelter was used (SMHI 1958), placed with the opening 1.5 m above ground.A comparison with data obtained with other gauges used in the forest (see below) showed that the results was usually within the reported range of observations (Nihlgård 1970), although occasionally they were 3% lower.
The throughfall was collected in 12 gauges randomly spread in a forested area of 1600 m 2 .The gauges consisted of 1-L polythene bottles with a plastic funnel with diameter 15 cm fitted in the lid.A filter of inert glass wool was placed in the bottom of the funnel to prevent material entering the bottle.To stabilise the gauges, they were placed on the ground inside metal tubes (diameter 16 cm).The top of the gauges was 30 cm above the ground.
The rain gauges were emptied after each rain event between December 2, 1966 and November 2, 1967.The volume was measured and the pH was determined in order to identify samples with impurities.The bottles were stored cool and then brought to the laboratory on 15 occasions during the measurement period for chemical analyses.

Chemical Analyses of Water
On arrival at the laboratory, the samples were pooled into four composite samples and kept cool (+4˚C).Analyses were carried out within one month as follows: pH was determined with a pH-meter, type Beckman N-2 (accuracy ±0.02).In calculating average pH, the pH values were first transformed to concentration of hydrogen ions and the average was converted back to pH (cf.Barrett & Brodin 1955).
Further analyses were performed on filtered water.The cations Na, K, Ca, Mg, Fe and Mn were determined on a Perkin-Elmer Atomic Absorption Spectrophotometer, with analytical error normally 1%, estimated from the standard curves.

Soil Sampling and Chemical Analyses
The soil sampling procedure and the methods applied for determination of exchangeable and total amounts of elements are described in full in Andersson, 1970b (pp. 141-146).Statistical analyses are also described in the same publication.

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

Biomass and Production of the Tree, Shrub and Field Layers
Biomass and primary production are essential as carrier substances of the elements treated in this paper, so the results presented by Andersson (1970b) for the Linnebjer nature reserve are summarised here.Tree biomass (B) of the mixed forest was estimated to be 238 t•ha −1 , with 201 aboveground and 38 t•ha −1 belowground (Table 1).

Element Amounts and Their Distribution in Biomass and in Production Fractions of the Tree, Shrub and Field Layers
The carbon content in the above-ground fraction of trees and shrubs was 114 t•ha −1 and that in the below-ground fraction was 19 t•ha −1 .A factor of 2 is often accepted for conversion of carbon to biomass.The nitrogen content was 0.81 t•ha −1 in above-ground tree and shrub biomass including also the field layer and 0.32 t•ha −1 in belowground biomass.For the finer fractions, a higher relative content was found.The field layer above and below ground contained minor amounts of elements (Table 2 and Table 3).In order to set the concentrations of elements and their distribution in the biomass fractions in the forest into perspective, a comparison was using beech forests along a fertility gradient in the same region (Table 4).The fertility can be described by the degree of base saturation of the soil.The Linnebjer forest had a similar degree (around 20%) as surrounding Lamium-type beech forests (Nihlgård & Lindgren 1977).Mercurialis-type beech forests have a high soil base saturation and Deschampsia-type a low soil base saturation.The Deschampsia-type forests also have a lower carbon content than nutrient-richer forests, while K, Ca and Mg are related to availabi- From: Oak (Andersson, 1970b); Beech (Nihlgård & Lindgren 1977).
lity in the form of exchangeable amounts in soil and in total biomass.Other elements such as P, Mn and Fe are dependent on the degree of solubility, which is regulated by soil pH.Sulfur to some extent reflects atmospheric deposition.

Amounts of Elements in Litterfall and Surface Litter
In order to analyse the circulation of elements, data on yearly litterfall (L) and the dynamics of the surface litter (SL) were used (Table 5).
The nitrogen content in the annual litterfall was 69 kg•ha −1 .The other elements were present in the following descending order: Ca, Si, Mg, K, S, Mn, P, Fe and Na, in amounts ranging from 41 to 0.9 kg•ha −1 •yr −1 .The surface litter comprised 6.0 t•ha −1 organic matter, or 2.8 t•ha −1 of carbon with a nitrogen content of 107 kg•ha −1 .The other elements occurred in the following descending order: Ca, Si, S, Mg, Mn, K, P, Fe and Na, in amounts ranging from 59 to 0.8 kg•ha −1 .Some readily mobile elements, such as K and Na, had a decomposition rate (k-value) of approx.50%, indicating that half the litterfall amount was turned over during one year.Other elements such as Si had a slow turnover rate of 14%, while the rest had intermediate values.

Amount of Elements in Precipitation, Throughfall and Interception
Input of elements was the result of precipitation measured in the open field (In), and of not measured dry deposited, aerosols (A) (Table 6 and see Figure 1).
Through fall (T) was measured and the difference between In and T (Diff) indicated leached (positive values) or absorbed elements (negative values) in the canopy.The Diff value also included possible dry deposited aerosols (A).What were possibly overlooked were elements from A that had been absorbed into the leaves.This can be the case e.g. for NO 3 and NH 4 ions.The amount of precipitation in the year of measurement ( 1967) was 644 mm.The amount captured by the canopy, interception, was 168 mm.The input of elements with rain was in decreasing order: Cl, S, N, Na, Ca, Mg, K, Mn and P in amounts ranging from 23 to 0.04 kg•ha −1 •yr −1 .The through fall contained in decreasing order: Cl, K, S, N, Na, Ca, Mg, Mn and P, in amounts ranging from 55 to 1.5 kg•ha −1 •yr −1 .The higher values show the most readily motile elements.
Special consideration needs to be given to deposition of sulfur and nitrogen.The open field received inputs of 11.0 kg S ha −1 •yr −1 and 9.4 kg N ha −1 •yr −1 .The through fall contained 34.3 S and 22.3 N kg•ha −1 •yr −1 .Over time, these amounts would contribute to acidification of the soil and one-sided fertilisation of the forest.The study area was located in a rural landscape not far from urban areas, which may explain the very high level of S, but also the unexpectedly high level of N. The levels observed were lower than those reported for the north-west of the region at the time (Nihlgård 1972).Since then, S deposition has strongly decreased and the Slevel in the open field today is approx.5 kg•ha −1 •yr −1 .The present day N deposition rate is approximately 10 kg•ha −1 •yr −1 (Pihl-Karlsson et al. 2012).

Soil Organic Matter and Exchangeable Mineral Content
The organic matter of the soil (SOM) was estimated to 288 t•ha −1 , with a carbon content of 108 t•ha −1 and a nitrogen content of 12 t•ha −1 (Table 7).The exchangeable (Exch) amounts in an extraction with ammonium acetate (Am-Ac) were in decreasing order: Ca, P, Mg, K, Na and Mn.

Discussion
Element cycling is currently a central theme in ecosystem ecology (Ågren & Andersson 2012, Andersson 1970a).The main results of the present study are summarised in Table 8.The yearly biomass increase indicates that at the time of measurement, the Linnebjer forest ecosystem is still in an aggradation or accumulation phase (Bormann & Likens 1979).Tree biomass increased yearly by 6.6 t•ha −1 or 3.1 t•ha −1 carbon.The other elements accumulated in the order: N > Ca > Mg > K > Mn > Si > P, in amounts ranging from 29 to 1.8 kg•ha −1 yearly.The litterfall returned elements to the soil.14% -53% was decomposed and returned to the ecosystem over the year.It can thus be assumed that the exchangeable pool of elements was slightly increasing.The atmospheric deposition of sulfur observed was sufficient to cause acidification of the forest soil.The nitrogen input should have had a fertilising effect, leading to increased forest growth.
The results reported deal with plant biomass, primary production and mineral cycling.Earlier the two first properties have been reported, but the mineral cycling not.As the results still are valid and can serve as reference material, it is considered essential to have them published.The cycling of carbon, nitrogen and sulfur are elements, which are affected by a changing climate and deposition.Repeated investigations make it possible to address questions like the forest as a source or sink for carbon.
The result of plant biomass and production are less affected by time as the methods are considered as more or less the same today.Chemical methods and techniques have changed more.The results related to chemistry and reported are, however, considered as "robust" and will allow being used for comparisons.Even if "modern" chemical methods are used, the final results are depending on the quality of the data from the estimation of plant biomass and production.
In a follow up paper, results from an investigation of sampling in vegetation and soil after fifty years will be reported.
Table 7. Distribution of elements in soil organic matter (SOM) and amount in an extraction with ammonium acetate (Am-Ac.Exch) in the soil profile of a mixed forest of Quercus robur and Corylus avellana in Linnebjer, Sweden.ver-failing study mate, Jan Ericsson, and staff at the Botanical Garden, Lund.Many measurements and samplings were performed by a local farmer.
The scientific work involved intense collaboration and exchanges over the years with Professor emeritus Bengt Nihlgård, Lund, who shared the final work.Scientifically, I am also indebted to Professor emeritus Göran Ågren, Uppsala, whose collaboration was an essential factor in finalising this work.
The Swedish Natural Science Research Council provided the main funding for the research.

Figure 1 .
Figure 1.Photo of the oak forest with Quercus robur and Corylus avellana also showing litter traps and a rain gauge.-Photo author May 1969.

Figure 2 .
Figure 2. Map of the Special area showing ecosystems and sample areas.Key: A. Quercus robur-Oxalis acerosella ecosystem; B. Quercus robur-Geumrivale ecosystem; C = Quercusrobur-Athyriumfilix-femina ecosystem, D = Clearing phase of ecosystem A and B; E = Filipendula ulmaria ecosystem; F = Carex flacca ecosystem and G = Carex caespitose ecosystem.In the sample area for ecosystem A the situation of 16 litter traps and 12 rain gauges are shown.

Figure 3 .
Figure 3. Schematic model of the turnover of organic matter and mineral cycling.For abbreviations see text.

Table 1 .
Plant biomass and yearly production (t•ha −1 dry weight at 85˚C) of the tree layer in a mixed woodland with Quercus robur and Corylus avellana.Linnebjer, Sweden.

Table 2 .
Plant biomass and yearly production (t•ha −1 dry weight at 85˚C) of the tree layer in a mixed woodland with Quercus robur and Corylus avellana.Linnebjer, Sweden.

Table 3 .
Mineral content (t•ha −1 dry weight at 85˚C) of a mixed forest of Quercus robur and Corylus avellana in Linnebjer, Sweden.

Table 4 .
Comparison of the element distribution in the biomass of a mixed forest of Quercus robur and Corylus avellana in Linnebjer, Sweden, with that in neighboring beech forests along a fertility gradient in southernmost province in Sweden.

Table 5 .
Content of organic matter and elements in litterfall (L) and surface litter (SL) and decomposition factor k (L/L + SL) of a mixed forest of Quercus robur and Corylus avellana in Linnebjer,Sweden, April 1966-March 1967.

Table 8 .
Distribution of organic matter and chemical elements in some important above-ground functional fractions and turnover characteristics of a mixed forest of Quercus robur and Corylus avellana in Linnebjer, Sweden.