Investigations on Spatial Heterogeneity of Humus Forms and Natural Regen ¬ eration of Larch ( Larix decidua Mill . ) and Swiss Stone Pine ( Pinus cembra L . ) in an Alpine Timberline Ecotone ( Upper Engadine , Central Alps , Switzerland )

Investigations on Spatial Heterogeneity of Humus Forms and Natural Regeneration of Larch (Larix decidua Mill.) and Swiss Stone Pine (Pinus cembra L.) in an Alpine Timberline Ecotone (Upper Engadine, Central Alps, Switzerland) The north-west facing study area in the Upper Enga¬ dine (Central Alps, Switzerland) encompasses the whole timberline ecotone between 2200 m and 2400 m a.s.l.. By dendroecological methods, clear differences concerning quantity and age structure of the natural regeneration of the main tree species Swiss stone pine (Pinus cembra L.) and larch (Larix decidua Mill.) at different sites could be detected.The differing site con¬ ditions, closely linked to the microtopography, deter¬ mine distribution and age structure of natural tree regeneration, as well as humus forms. On exposed sites characterised by Tenuic Humimors, an intensive nat¬ ural regeneration of, in particular, Swiss stone pine occurs. In depressions with a Rhizic Leptomoder, the density of tree regeneration is low and a high tree mortality could be observed. The results indicate that the site conditions most important for the natural regen¬ eration of trees in the timberline ecotone are reflected in the site-specific humus form. Zusammenfassung: Untersuchungen zur räumlichen Variabilität von Humusformen und natürlicher Rege¬ neration von Lärche (Larix decidua Mill.) und Arve (Pinus cembra L.) im Waldgrenzökoton (Oberenga¬ din, Zentralalpen, Schweiz) Auf einer nordwest-exponierten Untersuchungsfläche im Oberengadin, die den gesamten Bereich des Waldgrenzökotons zwischen 2200 m und 2400 m ü. NN umfasst, wurden mit Hilfe dendroökologischer Methoden deutliche Unterschiede hinsichtlich Anzahl und Altersstruktur der natürlichen Regeneration der waldbildenden Baumarten Arve (Pinus cembra L.) und Lärche (Larix decidua Mill.) festgestellt. Die in Abhängigkeit vom Mikrorelief kleinräumig wechseln90 Geographica Helvetica Jg. 57 2002/Heft 2 den Standortbedingungen bestimmen Verteilung und Mortalität des Baumwuchses sowie die Humusformen. Während auf exponierten Standorten mit der Humus¬ form Tenuic Humimor eine intensive Verjüngung ins¬ besondere von Pinus cembra L. zu beobachten ist, zeichnen sich Muldenlagen mit längerer Schneebe¬ deckung, auf denen ein Rhizic Leptomoder ausgebil¬ det ist, durch eine geringe Verjüngungsdichle und eine hohe Mortalität aus. Die Ergebnisse deuten darauf hin, dass die für die natürliche Regeneration der waldbil¬ denden Baumarten wesentlichen Standortbedingun¬ gen sich in den Humusformen widerspiegeln. Resume: Recherches sur la variabilite spatiale des formes d'humus et sur la regeneration du meieze (Larix decidua Mill.) ainsi que du Pinus cembra L. dans le domaine de Pecotone de bordure forestiere (Engadine superieure, Alpes centrales, Suisse) Sur une aire d'etude exposee au nord-ouest de l'Engadine superieure (dans les Alpes centrales suisses), qui englobe la totalite de l'ensemble de l'ecotone de la bordure forestiere etagee entre 2200 et 2400 m d'al¬ titude, il a ete possible d'observer, ä l'aide de methodes dendro-ecologiques, des differences notoires, en ce qui concerne le nombre et la structure d'äge, parmi les principaux arbres forestiers Pinus cembra L. et Larix decidua Mill. soumis ä la regeneration naturelle. L'evo¬ lution des rapports de dependance reciproque entre des aires exigues et le micro-relief determine la repar¬ tition et la mortalite des arbres, ainsi que les formes d'humus. Alors qu'il est notamment possible d'obser¬ ver un rajeunissement intense du Pinus cembra L. sur les aires ä humus Tenuic Humimor, les aires de depression ä couverture neigeuse prolongee sur lesquelles se developpe le Rhizic Leptomoder, se caracterisent par une faible regeneration et une mortalite elevee. Les resultats obtenus temoignent du fait que les principales conditions d'une regeneration naturelle de ces types d'arbres forestiers se refletent dans les carac¬ teristiques de 1'humus. Dr. Bettina Hiller, Institut für Didaktik der Geo¬ graphie, Westfälische Wilhelms-Universität Münster, Robert-Koch-Strasse 26, D-48149 Münster. Dr. Andreas Müterthies, Prof. Dr. Friedrich-Karl Holt¬ meier, Institut für Landschaftsökologie, Westfälische Wilhelms-Universität Münster, Robert-Koch-Strasse 26, D-48149 Münster. Prof. Dr. Gabriele Broll, Abteilung für Geoökologie, ISPA, Hochschule Vechta, Postfach 1553, D-49364 Vechta. e-mail: hillerb@uni-muenster.de muetert@uni-muenster.de Holtmei@uni-muenster.de gbroll@ispa.uni-vechta.de Manuskripteingang/received/manuscrit entre le 15.1.2002 Annahme zum Druck/accepted for publication/accepte pour Timpression: 23.5.2002

Investigations on Spatial Heterogeneity of Humus Forms and Natural Regen¬ eration of Larch (Larix decidua Mill.) and Swiss Stone Pine (Pinus cembra L.) in an Alpine Timberline Ecotone (Upper Engadine, Central Alps, Switzerland) Bettina Hiller, Andreas Müterthies, Friedrich-Karl Holtmeier, Münster, Gabriele Broll, Vechta 1 Introduction and objective The timberline ecotone, a transitional belt between the subalpine forest and the alpine zone, is characterised by a patchwork of different microsites (e.g.Holt¬ meier & Broll 1992, Butler & Domergue 1997).
The microtopography of the ecotone strongly affects most site factors, for example, snow cover, soil moisture and soil temperature, leading to different Veg¬ etation and soil properties between the microsites (Braun-Blanquet et al. 1954).Both the site condi¬ tions and the general modes of seed dispersal (windmediated or by the European nutcracker (Nucifraga caryocatactes caryocatactes)) cause a heterogeneous regeneration of the main tree species reflected in the distribution of seedlings and saplings in the ecotone.
The aim of this study is to analyse both the impact of the different site factors and their interaction on the altitudinal differentiation of the natural regeneration of the main tree species in the timberline ecotone.The humus forms.influenced by Vegetation and soil organisms, as well as abiotic factors, like relief and cli¬ mate, are investigated as to their suitability as indica¬ tors for potential successful natural regeneration of European larch (Larix decidua Mill.) and Swiss stone pine (Pinus cembra L.). 2

Study area
The study area is the north-west facing slope of the Piz da Staz (2847 m a.s.l.) in the Upper Engadine, Central Alps, Switzerland (Fig. l).The study area has a slightly Continental climate, characterised by rela¬ tively low preeipitation and high solar radiation.Par¬ ticularly due to alpine pasturage, which was stopped in the 50s, the timberline of the study area retreated to 2200 m a.s.l..As reflected in the snow patterns of the timberline ecotone (located at about 2200 -2400 m a.s.l.), the prevailing microclimatic conditions are strongly influenced by the locally varying micro¬ topography.Talus deposits of gabbrodiorite.diorite and essexite make up the parent material (Staub 1946).The topography is characterised by rocky outcrops, knolls and small ridges alternating with depres- sions.Leptosols, Cambisols and Podzols are common (Müller 1983).In depressions, pedogenesis is often influenced by the accumulation of fine mineral and organic material.Swiss stone pine and European larch form the timberline.These trees also dominate the subalpine Larch-Stone pine forest (Larici-Pinetum cembrae) (Keller et al. 1998).In the timberline eco¬ tone the locally varying site conditions cause a mosaiclike Vegetation pattern.Dwarf shrubs.like Vaccinium gaultherioides and Vaccinium myrtillus, are common.Loiseleuria procumbens and various liehen species are typical of wind-exposed locations.As the snow cover in depressions tends to last longer than on the surrounding area and due to the moist to wet con¬ ditions during the growing season, the Vegetation of these areas is mostly dominated by sedges, grasses and 3 Material and methods The study site, covering an area of 6 ha, extends over the whole timberline ecotone between 2200 m and 2400 m a.s.l..The data presented in this paper were collected from six representative plots of 20 x 20 m.Plot 1, located at the treeline in the central part of the eco¬ tone at 2300 m a.s.l., is chosen as an example.In 1997 and 1998, the following factors were mapped at the study site using a 10 m grid and recorded in detail on the six plots: microtopography, snow melt-out, Veg¬ etation, soil conditions and humus forms.Additionally, differences in soil conditions and humus forms between exposed sites and depressions were investi¬ gated on the basis of transects (20 m long).On the plots, microsites representing similar characteristics were defined.Microsite A represents an exposed site with melt-out before 22.05.1998.microsite B a depression with melt-out after 28.05.1998.The Vegetation cover was estimated as a percentage of surface for three different categories at each microsite: vascular plants, lichens and mosses.Dendrochronological meth¬ ods were used to determine the age and tree ring width of the examined trees.Samples were taken from trees with a diameter larger than five centimetres by an increment borer.For trees with a smaller diameter either an increment puncher was used (Forster et al.   2000) or dises were cut.Special attention was given to taking the dises and cores as close to the stem basis as possible.To establish missing and false tree rings and to date the years of germination and death of the examined trees, crossdating (Douglass 1941)     Reproduced with permission of the Federal Office of Topography (BA002176)   done with TSAP (Time Series Analysis Program).Furthermore, height, basal area, number of whorls, growth form and damage were recorded for all exam¬ ined trees.The humus forms of the study site were described and classified according to both Green et   al. (1993) and the Swiss soil Classification System (FAP 1992).The master organic horizons of the Classifica¬ tion according to Green et al. (1993) correspond with the Swiss or German Classification System for humus forms (FAP 1992, AK Standortskartierung 1996).
Samples were taken from both the organic layer and the mineral topsoil from each perceived horizon.In situations where no horizon could be identified, sam¬ ples were taken at 2 cm depth intervals.The soil tem¬ peratures were recorded with dataloggers (StowAway TidbiT) at a depth of 2.5 cm on two microsites from 26 May 1998 to 11 October 1998.

Results
The duration of snow cover on convex sites is usually shorter than on concave topography (Fig. 2 and 3).
These differences are clearly reflected in the plant cover.Microsite A is dominated by dwarf-shrubs, such as Vaccinium gaultherioides and Loiseleuria procumbens, and lichens.The plant density is low, sometimes the bare mineral soil is exposed.At microsite B, the Vegetation consists of sedges, grasses, mosses and herbs (mostly Trichophorum caespitosum and Carex fusca), indicating temporary moist site conditions.The plant density on these sites is markedly higher than on the more wind-exposed sites (Table 1).
Microtopography and snow cover also influence the soil conditions and humus forms of these two micro- sites.The humus form of microsite A can be defined as a Tenuic Humimor (Green et al. 1993) or as a «typischer feinhumusreicher Rohhumus» (FAP 1992) with a very thin organic layer (Fig. 4).The L-layer is composed of Vaccinium and Loiseleuria leaves.A few fungal mycelia were observed.Faunal droppings could not be found.Tire H-horizon (about 2 cm thick) con- tains finely dispersed organic matter and mineral particles.The underlying A-horizon is affected by pod- zolization (Fig. 4).The humus form of microsite B Vegetationsbedeckung und Temperaturen in 2.5 cm Tiefe unterhalb der Geländeoberfläche (Mai bis Okiober 1998) auf den Kleinstandorten A und B auf der Aufnahmefläche 1 Couverture vegetale et temperatures ä 2,5 cm de profondeur (de mai ä octobre 1998) sur les microsites A et B de  Melt-out 1998 after 28.05.Anzahl von «Pinus cembra L.» und der vom Tannenhäher geöffneten Samendepots pro m2 in Abhängigkeit vom Ausaperungszeitpunkt auf der Aufnahmefläche 1 Nombre de «Pinus cembra L.» et de elepöts de semences/m-, ouvertes au casse-noisette (aire d'observation 1) lors de la fönte de la couverlure neigeuse can be described as a Rhizic Leptomoder (Green et  al. 1993) or as a «typischer feinhumusreicher Moder» (FAP 1992).Being 9 to 11 cm deep, the thickness of the organic layer of microsite B exceeds that of microsite A.The L-layer contains mainly leaves and grass stems.Only a few dwarf shrub leaves are to be found.The litter is «non-compact matted».The underlying H-horizons contain predominantly finely dispersed organic matter.The numerous fine and very fine roots with a mat-like structure in the first 2 cm of the H-horizon are striking.The root density decreases in the lower hori¬ zons (Fig. 4).Results of soil temperature monitoring at a depth of 2.5 cm between May and October 1998 likewise reflect differences between microsites A and B. The average temperature at microsite A is 1 °C higher than at microsite B.Not only does microsite A have the highest temperatures, the lowest tempera¬ tures were also recorded on this site (Table 1).
The density of young stone pines and the number of seed Caches established by the European nutcracker correspond to the duration of the snow cover in a given area (Fig. 5).Microsite A has a density of 0.55 seed Caches per Square meter, whereas at microsite B the density is only 0.07 Caches per Square meter.The number of Swiss stone pines germinating from these Caches amounts to 1.3 per Square meter for microsite A and 0.15 for microsite B. The correlation coefficient (Spearman) between the date of melt-out and the number of Swiss stone pines is 0.5.The same holds true for the correlation between the date of melt-out and the number of seed Caches.The age structure of Swiss stone pines on plots 1-6 indicates that the survival rate of seedlings decreases with increasing duration of the snow cover (Fig. 6).

Discussion
The microtopography is an important factor Control¬ ling spatial distribution and growth of natural tree regeneration in the alpine timberline ecotone.For -stand from plots 1 -6 in relation to the date of melt-out (n 192) Altersstruktur des «Pinus cembra L.»-Bestandes der Standorte 1 -6 in Abhängigkeit vom Ausaperungszeitpunkt Structure d'äge du «Pinus cembra L.» (aires d'observation 1 -6), en relation avec n 192 (lors de la fönte de la couverture neigeuse) example, the distribution of stone pine-seed Caches is strongly influenced by the microtopography.The larger number of Caches observed on microsite A (Fig. 2 and Fig. 5) may be explained by the nutcracker's seed cach- ing behaviour.The bird prefers for different reasons exposed sites with little or no snow cover (Holtmeier   1966, Mattes 1982, Kajimoto et al. 1998).It appears obvious that the nutcracker would need to make the least effort looking for seed at sites with little or no snow, thus saving energy.The bird's success rate in find¬ ing seed is also higher at such sites (Mattes 1982), not only because of little snow cover but also due to better orientation.It may also be easier for the nutcracker to establish seed Caches on microsite A because of the lower plant cover alternating with patches of bare mineral soil (Table l).The higher temperatures regis- tered at this site probably encourage a greater germination rate of conifer seeds than at microsite B (Table 1).On the other hand, snow accumulation at micro¬ site B protects young trees and seedlings from cli¬ matic injuries (frost, desiccation, ice particle abrasion) during the winter months.Although Swiss stone pines are comparatively frost resistant, younger pines are susceptible to frost during the growing season (Tran¬ quillini 1979).The duration of the snow cover can also negatively affect the number of stone pines in an area.Phacidium infestans, a fungus infecting the nee- dles and shoots of pines covered too long by snow (Roll-Hansen 1989), is responsible for the greater mortality of pines on microsite B.
The accumulation ofthe wind-dispersed seeds of Euro¬ pean larch may be higher in surface depressions due to lower wind speeds, but the ability of this species to germinate successfully at such sites is low.One reason for the low germination rate is the dense plant cover which prevents the light larch seeds from reaching a suitable seed bed.Usually, larch seeds need bare min¬ eral soil for germination (Auer 1947).Possibly for this reason the number of young larches at the study site is low, particularly as the density of Vegetation increased after pasture abandonment at the study site in the early fifties.  1961Bochter et al. 1983), as were also found on the study site in the Upper Engadine.The microtopogra¬ phy is an important factor for the humus form devel¬ opment in the timberline ecotone (Lüscher 1991,  Bednorz et al. 2000).Due to particular site micro¬ topography and snow cover duration, the properties of the organic layers and the topsoil at microsites A and B on plot 1 differ, too.The temperatures during the growing season at a depth of 2.5 cm are higher on microsite A than on microsite B. The temperatures are influenced by microtopography, Vegetation and the overlying organic or mineral material.Under the scat- tered dwarf shrubs and lichens of microsite A, the Upper 2.5 cm of soil contain slightly humified organic material, while at microsite B, the dense Vegetation, mostly sedges, grasses and mosses, leads to a high con¬ tent of well humified organic matter and abundant fine roots.The differences in the morphology of the organic layers of microsite A and B also influence the soil moisture conditions.The well humified organic material of microsite B enhances the water holding capacity of the soil, resulting in moist to wet soil con¬ ditions during the growing season.Moreover, during snow melt the snow-rich microsite B receives addi¬ tional melt water from adjacent slopes.The temporary moist to wet soil conditions during the growing season are also reflected in the Vegetation by the dominance of Carex fusca and Trichophorum caespitosum.The conditions at microsite A are different as dryness may occur in summer caused by a well-drained Substrate and the lower water holding capacity of the slightly humified organic material.A Vegetation dominated by dwarf shrubs favours an increase ofthe soil acidity and the development of Mor humus forms, thus leading to the occurrence of podzolization processes as indicated by an Ahe horizon.However, the Mor humus forms at the study site in the timberline ecotone have a thinner H and F horizon (Tenuic Mor or «Hagerhumus» after AK Standortskartierung 1996) than the humus forms of the montane belt.In particular, the translocation of litter and organic fine material by wind and water prevents the development of thick Mor layers at these exposed sites.Both the humus form and the occurrence of podzolization in the mineral topsoil indi¬ cates that the content of plant available nutrients is probably lower at microsite A than at microsite B. As the natural regeneration of Swiss stone pine is greater on the nutrient poor microsite A, it may be presumed that the nutrient content of a soil is less important than the other dominating factors: date of melt-out, Vegeta¬ tion cover and soil temperature. 6

Conclusions
As has been shown, the establishment, distribution pattern and development of larch and stone pine and the occurrence of different humus forms are control¬ led by microtopography influencing other site factors such as snow cover duration.soil temperature and soil moisture.At an exposed site, as represented by micro¬ site A, the Vegetation, dominated by dwarf shrubs and lichens, and an early snow melt lead to the formation of a Mor humus form.The abundant natural regen¬ eration of Swiss stone pine at this site may also be attributed to these factors.Due to the dense sedge and moss Vegetation of a relief depression, represented by microsite B, and the closely related late melt-out cor¬ responding with moist soil conditions, a Moder humus form developed on microsite B. These factors are also responsible for the low natural regeneration of Swiss stone pine and a high tree mortality at this site.There¬ fore, humus forms probably correspond with the site quality for natural regeneration of seedlings and sap- lings.Further detailed investigations of chemical and physical soil properties should help to better define site quality for natural regeneration of trees at the timber¬ line.The Classification System for humus forms accord¬ ing to Green et al. (1993) is best suited for describing and classifying the humus forms of mountain ecosys¬ tems in general (Bednorz et al. 2000).The Classifica¬ tion can be used in the alpine belt where Moder humus forms are very common.It may also be applied to closed forests where Mor humus forms prevail.How¬ ever, a more detailed differentiation of the humus forms, especially of the Mor order, in the timberline ecotone is necessary due to the extreme patchiness of site conditions.
The north-west facing study area in the Upper Enga¬ dine (Central Alps, Switzerland) encompasses the whole timberline ecotone between 2200 m and 2400 m a.s.l..By dendroecological methods, clear differences concerning quantity and age structure of the natural regeneration of the main tree species Swiss stone pine (Pinus cembra L.) and larch (Larix decidua Mill.) at different sites could be detected.The differing site con¬ ditions, closely linked to the microtopography, deter¬ mine distribution and age structure of natural tree regeneration, as well as humus forms.On exposed sites characterised by Tenuic Humimors, an intensive nat¬ ural regeneration of, in particular, Swiss stone pine occurs.In depressions with a Rhizic Leptomoder, the density of tree regeneration is low and a high tree mortality could be observed.The results indicate that the site conditions most important for the natural regen¬ eration of trees in the timberline ecotone are reflected in the site-specific humus form.

Fig. 4 :
Fig.4: Humus forms on the microsites A and B of plot 1 Humusformen der Kleinstandorte A und B auf der Aufnahmefläche 1 Formes d'humus des microsites A et B de Taire d'observation 1