PRODUCTION AND TURNOVER OF ORGANIC MATTER IN THREE SOUTHERN EUROPEAN FAGUS SYLVATICA

* IRNA-CSIC, Cordel de Merinas, 37071 Salamanca (Spain), ce.: ignac@gugu.usal.es ** Facoltà di Scienze Matematiche, Fisiche e Naturali, Universita di Catania, 19via A. Longo, 95125 Catania (Italy). Corso di Laurea in Scienze Forestall ed Ambientali, Universita di Palermo, 23 via Roma, 92010 Bivona (Italy) ***CNRS-CEFE, BP 5051, 34033 Montpellier Cedex (France) ****Servicio de Medio Ambiente. Junta de Castilla y León. Villar y Macías 1, Salamanca (Spain)

SUMMARY.-Above-ground biomass, litterfall and litter accumulation and decomposition at the soil surface were studied within three Mediterranean beech forests from Italy, France and Spain in order to better understand the recycling of elements associated with the turnover organic matter Above-ground tree biomass amounted to 131.9 Mg ha'^ at Etna (Italy), 134.2 Mg ha'' at Sierra de la Demanda (Spain) and 223.9 Mg ha '' at Mont Lozère (France).
The highest amount of total litterfall was observed at Sierra de la Demanda (4.7 Mg ha-' year'), followed by the Mont Lozère (4.4 Mg ha' year' ) and Etna (3.9 Mg ha' year').
Total organic matter accumulated on the soil surface in the three beech forests amounted to 25.8 Mg ha'' at Mont Lozère,14.4Mg ha'' at Sierra de la Demanda and 12.6 Mg ha'' at Etna.The relative proportions of leaf litter versus total litter were nearly the same in the Etna and Sierra de la Demanda forests (72 -70%), and close to these values for Mont Lozère (65%).
All the studied Mediterranean Fagus sylvatica stands appeared very similar concerning the organic matter distribution and fluxes, even if local climate and soil differences can be noticed.
Toutes les forêts de Fagus sylvatica dans le climat Méditerranéen sont très similaires par rapport à la distribution et aux flux, malgré que les conditions de climat et de sol soient différentes.

Introduction
The biogeochemical cycle of organic matter and mineral elements plays a key role in the relationships between the soil, the vegetation and the surrounding environment.This cycle is one of the crucial ecological phenomena in natural biocenosis and, in particular, in forest ecosystems (RAPP, 1969).The most important contribution to the soil humus occurs through above-ground and root litter (GOSZ et al, 1976).Aboveground litter plays a fundamental role in the nutrient turnover and in the transfer of energy between plants and soil, being the source of the nutrient accumulated in the uppermost layers of the soil.This is particulary important in the nutrient budgets of forest ecosystems on nutrient-poor soils, where the vegetation depends in large part on the recycling of the nutrients contained in the plant detritus (SINGH, 1978, VOGT et al. 1986).
In forest ecosystem, litter production is mainly expressed as a massive contribution of dead organic matter that accumulates on the ground (MANGENOT & TOUTAIN, 1980).This accumulated leaf litter on the soil surface, together with the contribution made by root decomposition (McCLUGHERTY et al, 1982), represents the basic source of energy, C, N, P, and other bioelements for the participating microflora and mesofauna of soil, as well as a quantity of easily available nutrients (RAPP & LEONARDI, 1988).
The aim of the present work was to determine and compare distributions of above-ground living biomass and detrital biomass: litterfall and litter accumulation, in three Fagus sylvatica forests, under a Mediterranean climate, in Spain, Italy and France, differing nevertheless by local climate, soil characteristics, site history, topographical position, stand age and stand structure.

Study sites
Three Fagus sylvatica stands were selected in the Etna volcano (Italy), Sierra de la Demanda (Spain) and Mont Lozère (France), to compare different stand characteristics in a Mediterranean area (Figure 1).
All stands are located in mountain-Mediterranean climate involving cold winters and hot dry summers, with Mediterranean, maritime and mountain climatic influences.
The climate is meso-Mediterranean in the Etna volcano, with a mean annual temperature of 10°C and a mean annual rainfall of 660 mm.
In Sierra de la Demanda, the climate is also meso-Mediterranean, becoming sub-Mediterranean with increasing elevation.Mean annual temperature and precipitation are 12.4°C and 895 mm respectively.The average length of the summer dry period is 2 months per year (July and August) and the length of the cold period (monthly average temperature less than 7°C) is 6 months per year.
In Mont Lozère, the climate is mountain-Mediterranean, with a mean annual temperature of 6.5°C.The mean monthly temperature limits were PIRINEOS 153-154 Figure 1.Situation of research sites on the Mediterranean area.Situación de las parcelas de experimentación en el área mediterránea.
obtained in August 1991 (+17.4°C) and in December 1990 (-2.3°C).There were more than 150 days with frost.Mean annual rainfall was 1431 mm, with a very irregular distribution along the year, high rainfall levels in the autumn and spring, and long summer droughts.Snowfall contribution varied yearly, between 2.9% and 14.7% of overall precipitation.
The beech forest of the Etna volcano region is a naturally forest of over 45 years old, situated between 1450 m and 1750 m elevation, with a density of 2124 stems ha"^ (LEONARDI et al, 1996).The diameter distribution of trees is shown in Figure 2. The mean height ranged from 9 to 13 m, and the basal area was 10.1 m? ha-i (Table 1).
In Sierra de la Demanda, the beech stand is also a naturally forest, 50 years old, located at 1000 m over sea level, with a density of 523 trees ha'^ and tree Table 1.Stand characteristics of the three Fagus sylvatica forests.Características de las tres parcelas de hayedo.
diameters ranging from 13 to 63 cm.The diameter distribution classes are shown in Figure 2. The mean height ranged from 20 to 22 m and the basal area was 25.7 m2 ha'^ (SANTA REGINA et al, 1996).
In Mont Lozère, the beech forest is nowadays naturally regenerated, after being planted at the beginning of the century.Uneven-aged between 50 and 80 years, the forest is located 1200 m above sea level.The density of this forest is 4270 trees ha'^.The diameter distribution classes are also shown in Figure 2. The mean height ranged from 7 to 8 m and the basal area was 43.0 m^ ha"^ (HANCHI, 1994).
The soil from the Etna beech site is developed on basaltic lava and classified as Humic Andosol.The soil depth ranged from 20 to 40 cm.HOFMANN (1960) indicates that the soil is poor in clay content with high sand (74%) and lapilli (13%) content, composed only by a dark A horizon, covering the lava substratum.The pH (H2O) does not vary with depth and is around 6.
At Sierra de la Demanda the soil is classified as Humic Acrisol.There are a wide range of soil depths, the clay content increasing with depth and the soil is developed on a paleozoic lithology.The pH (H^O) is acidic, varying from 5.8 at the surface to 5.4 at 100 cm depth.The soil has a high organic matter content decreasing with depth from 15.9% to 3.3%, with a C/N relation around 20.0 at surface and 4.0 at depth.
The soil of Mont Lozère is a mountain ranker, developed on porphiroid granite (VANNIER, 1992).The soil depth ranges from 40 to 70 cm.The pH (H2O) is acidic, varying from 4.6 at surface to 5.4 at 60 cm depth.The soil from Mont Lozère has a high organic matter content, with organic matter varying also, between 5 and 2 % with increasing depth.

Biomass determination
We used allometric relationships between the biomass and the diameter at breast height (DBH) of trees to estimate stand biomass.The DBH of all trees (c) Consejo Superior de Investigaciones Científicas Licencia Creative Commons 3.0 España (by-nc) http://pirineos.revistas.csic.esPIRINEOS 153-154 in plots of 1 hectare area was measured in each experimental site (1992) and diameter distributions calculated (Figure 2).Thirty representative trees of the different DBH classes were felled (13 in Italy, 7 in Spain and 10 in France) to establish above-ground biomass equations.
Each harvested tree was divided into stem, branches and leaves.Trunk sections, according to height (0-1.3 m, 1.3 -3 m, 3-5 m, ^-1 m, 7-9 m and so on), were weighed in the field along with total branch mass and individual branches.Subsamples for trunk, branch and leaf were brought to the laboratory for measurement of moisture content after drying to constant weight at 80°C.
For each tree, trunk, branch and leaf dry mass and total dry biomass were calculated and correlated with DBH using regression analysis.Various regression equations, adjusted for the 30 sample trees, indicated the following determination coefficients: 0.679 for logarithmic regressions, 0.801 for exponential regressions, 0.968 for linear regressions and 0.984 for power regressions.The power regression equations: Biomass (Kg) = a (DBH^^)*", having the highest determination coefficients were used to estimate the total biomass of the different above-ground tree compartments.
Yearly wood production was estimated by dividing, for each stand, the above-ground perennial biomass by the mean age of the trees.This implies some approximations because of the doubts about the age of all trees from each stand, variations in production over the lifespan of the stands and death and disappearance of some trees during the same period.

Litter fall
Thirty litter traps, with 0.25 m^ interception area, were distributed randomly at the Spanish site and 28 at the French and Italian sites during a three years period (1990)(1991)(1992).The litter traps were emptied monthly during autumn and the collected material was subdivided into leaves, reproductive material, wood and undetermined material, then oven dried to constant weight at 80° C and weighed.The results are expressed on a stand surface area (ha).

Litter layer
Accumulated total litter on the soil surface was estimated by collecting the • organic layer (Aoo) from 15 samples (0.5 x 0.5 m) in Spain, 16 in France and 12 samples (0.25 x 0.25 m) in Italy, randomly chosen in all three stands.The inorganic Aj^ layer was not included in these samples.Total litter was then dried and weighed.
The decomposition coefficient K of JENNY et al. (1949) was used to assess litter dynamics.It describes yearly litter decomposition in relation to humus type, above-ground litter and yearly litterfall.K is defined by: K = A / (A + F) where A is the amount of litter returning annually to the soil, and F is the litter accumulated on the soil surface, before the period of massive litter fall (SANTA REGINA & TARAZÓN A, 1995).

Above-ground biomass o/Fagus sylvatica trees
Power regression equations, calculated from the different samples of trees of the study are given in Table 2.There were generally strong correlations  for each tree compartment and for the entire tree for the three stands studied.

Relación entre la biomasa aérea (y) y el DBH (x) para cada compartimento del árbol o para el árbol entero, todo ello en las tres parcelas estudiadas.
between biomass of the various components (trunks, branches, leaves) and DBH.As shown in Figure 3, the allometric relations between total biomass and DBH were very close for the three stands.This allowed us to calculate another regression, pooling the three stands.This last regression is not statistically different from that calculated for the Mont Lozère site.From these results pooled regressions were also calculated for the aboveground biomass of Fagus sylvatica tree components.The relative proportions of foliage, branches and trunks in relation to DBH are shown in Figure 4.
Above-ground tree biomass amounted to 131.9 Mg ha"^ at Etna, 134.2 Mg ha"^ at Sierra de la Demanda and 223.2 Mg ha'^ at Mont Lozère (Table 3).Although the amounts were very different between stands, especially between Mont Lozère and the two other sites, the proportions for the different tree compartments (leaves, branches and trunks) were very similar.
. Foliage represented a relatively low proportion of total above-ground tree biomass.In,the Etna stand, foliage amounted to 3.3% (Table 3), very similar to the level in the Sierra de la Demanda stand: 3.2%.In the Mont Lozère stand although leaf biomass was higher, the proportion of leaves to above-ground biomass amounted to only 2.1%.Wood amounted to 97% of the above-ground biomass, and approximately 80% of this wood was from trunks and 20% from branches, with the following values for each stand: 76% trunk biomass and 20.8% branch biomass at Etna, and respectively 75% and 21.9% at Sierra de la Demanda, and 76.2% and 21.4% at Mont Lozère.
Mean yearly wood production in the 3 stands amounted respectively to 2.6, 2.8 and 4.4 Mg ha'^ year^ at Sierra de la Demanda, Etna and Mont Lozère (Figure 5).

Litterfall
Table 4 summarizes annual litterfall for the three stands.The highest amount of total litterfall was observed at Sierra de la Demanda, followed by the Mont Lozère and Etna sites.The differences resulted more from wood, inflorescence and fruit litter, than from leaf litterfall.Leaf litterfall amounts were the highest with similar values at Etna and Mont Lozère.
The differences appearing between the estimated leaf biomass and the leaf litterfall are mostly related to the date of biomass sampling.Canopy leaf mass varies during the season.If the biomass is estimated in summer, at the peak of leaf growth, this could explain the differences with leaf litter amounts.In addition, leaf litter was sampled only from September to December, under estimating some possible earlier leaf litterfall.

Litter accumulation on the soil
Total organic matter accumulated on the soil surface in the three beech forests amounted to 25.8 Mg ha'^ at Mont Lozère, 14.4 Mg ha"^ at Sierra de la Demanda and 12.6 Mg ha"^ at Etna (Table 5).Although total organic matter accumulation varied markedly between the three stands, the relative proportions of leaf litter versus total litter accumulated were nearly the same in the Etna and Sierra de la Demanda forests (72 -70%), and close to the values in the Mont Lozère forest (65%).* Measured for the whole lifespan of the stands./Merf¡rf«en el total de vida de las tres parcelas.

Discussion
Table 6.Above-ground perennial productivity for the three Fagus sylvatica stands.Productividad de biomasa perenne en las tres parcelas de hayedo.
estimation of the root biomass, calculated from values indicated by other authors and summarized in Table 7.The mean value of 16% of above ground biomass was used for the estimation of root biomass is the three studied stands.
Mean annual wood production amounted to 2.6 and 4.4 Mg ha'^.At the Etna site, the 2.8 Mg ha'^ wood production determined for the total lifespan of the stand corresponded to a current production of 6.2 Mg ha'V observed on the basis of the increase in DBH of all trees of the stand over 1 year (LEONARDI et al, 1996).At Mont Lozère, HANCHI (1994) found closer amounts with respectively 4.1 Mg ha'^ with the direct method (tree DBH increase during one year) and a mean value of 4.4 Mg ha'^ for the whole lifespan of the stand.
In all cases, mean productivity, expressed as percent of above ground perennial biomass, was similar for all tree Fagus sylvatica stands, with annual above ground perennial productivity of 2.0 to 2.2 %.
The 24.5% and 14.6% decomposition rates at Sierra de la Demanda and Mont Lozère, determined in the present study according to JENNY et al. (1949), could be compared to the decomposition rates of 29% at Sierra de la Demanda reported by SANTA REGINA & TARAZÓN A, (1995) and to'the 21.2% found at Mont Lozère (HANCHI, 1994) , 1996SANTA REGINA et al, 1996BRUNO et al, 1973HANCHI, 1994ULRICH et al, 1974ibid. HALLBACKEN, 1992ibid. ULRICH et al, 1974NIHLGARD, 1972DUVIGNEAUD et al, 1977 Table 7.Comparison of biomass and productivity in several Fagus sylvatica stands.Net primary production = * mean annual above-ground perennial increment + litter fall.** only yearly increase in above-ground perennial material.*** yearly increase in above-ground perennial material + litter fall.Comparación de la biomasa y la productividad en varias zonas de Fagus sylvatica.
In other beech forests, GLOAGUEN & TOUFFET (1980)  The Etna and Sierra de la Demanda stands were very similar in terms of amounts of organic matter determined for the different tree compartments and in terms of fluxes.The Mont Lozère stand had higher biomass values and higher litter accumulation on the soil surface, while some fluxes, namely to litterfall and yearly litter decomposition, were close to the levels obtained at Sierra de la Demanda.
To compare the three sites, we calculated the different fluxes and organic matter compartments as a percentage of the above-ground perennial biomass for each of the three stands.The same picture was found between the Etna and the Sierra de la Demanda stands, with annual leaf production of 3.3% and 3.2% of the perennial biomass, a mean annual increase of 2.2% and 2.0% of the perennial biomass and 3.0% and 3.6% litterfall, respectively.Annual decomposition of the litter layer amounted to 2.3% and 2.8% of the perennial above-ground biomass.
At Mont Lozère, the fluxes were lower, in relation to the higher aboveground perennial biomass: leaf production: litterfall: litter decomposition: 2.1% 2.0% 1.7%Only the mean annual increment of wood biomass (2.0%) was the same as in the two other stands.Generally three factors could be responsible for these differences: stand age, soil and climate.The similarity between Etna and Sierra de la Demanda was probably more related to the age of the stands than to soil and climate patterns, which were very different.The differences between the Mont Lozère and the two other beech sites concern mainly leaf production and organic matter turnover.
In comparison to other Fagus sylvatica forests (Table 7), total above-ground biomass of the studied stands was closely correlated with the age of the trees, but such comparisons are hazardous.Indeed, total above-ground biomass varies also according to the geographical location, forest type, forest structure and degree of management or disturbance (BROWN et al, 1989).The three Mediterranean sites had very high amounts of litterfall, as compared to other Fagus sylvatica forests.Only the 130 year old beech forest studied by NIHLGARD (1972), showed a higher litterfall.However the ratio of the annual litterfall to above-ground biomass was lower than that determined in the three presently studied stands, indicating again a clear relation between stand age and leaf production.

Conclusions
. (' There were similar levels (amounts and percentages) of organic matter production, accumulation and turnover in the three beech stands studied in France, Italy and Spain.
' ^ However the stands in Italy and Spain showed more similarities between them than with the French site.This latter site had higher amounts of aboveground biomes and litter, lower amounts for the main vegetation / soil fluxes: litterfall and litter decomposition, due perhaps, to its higher precipitation and mean annual temperature, or its different stand management and soil properties.
But all the three studied Mediterranean Fagus sylvatica stands appeared very similar concerning the organic matter distribution and fluxes, even if local climate and soil differences can be marked.

Figure 3 .Figure 4 .
Figure 3. Regression curves for biomass (Kg) in relation to DBH (cm)for the three stands and mean curve for Fagus sylvatica.Curvas de regresión para la biomasa (Kg) en relación con su DBH (cm), para las tres parcelas y curva media para Fagus sylvatica.

Figure 5
Figure5summarises the above-ground organic matter compartments and fluxes at the three studied sites; the amounts of annual wood production and annual litter decomposition have been added to this figure as well as an (c) Consejo Superior de Investigaciones Científicas Licencia Creative Commons 3.0 España (by-nc) http://pirineos.revistas.csic.esPIRINEOS 153-154 Consejo Superior de Investigaciones Científicas Licencia Creative Commons 3.0 España (by-nc) http://pirineos.revistas.csic.es