Diversity and tree neighborhood effects on the growth dynamics of European beech and the stand seed bank in temperate broad-leaved forests of variable tree diversity

Abstract

Among the recent studies investigating species diversity s effect on ecosystem functioning, only few examine long-term changes. My study addresses this question based on seed bank analysis and dendrochronology in temperate deciduous forest stands in Hainich National Park (Thuringia, Germany). I focused on three main hypotheses i) species identity of the neighbor trees and their functional traits can significantly influence growth dynamics of Fagus sylvatica L. and the soil seed bank, while species diversity of the neighborhood is of no significance, ii) neighborhood diversity alters trees capacity of reaction to disturbances, and iii) there are processes altering tree species diversity, growth patterns of F. sylvatica and the soil seed bank at the same time.The selected 12 forest stands, situated on similar bedrock, represent a gradient in tree species diversity. Samples were taken from monospecific beech stands (DL1), medium-diverse forests with beech, ash, and lime (DL2), and highly diverse stands dominated by beech, ash, lime, maple, and hornbeam (DL3). I investigated germinating seeds in three horizons (0-5cm, 5-10cm, 10-20cm) of the soil seed bank and determined seed density, composition and functional types of species. Furthermore, structural parameters of target Fagus trees and their neighbors were analyzed. Neighborhoods of target trees were grouped according to the relative importance of intraspecific competition by Fagus: from purely conspecific neighborhoods (Fagus100) to neighborhoods with mainly conspecific neighbors and few allospecific competitors (Fagus70-99) to neighborhoods with more than 30% of the canopy s influence sphere occupied by allospecific trees (Fagus<70). I examined tree-ring width and growth response to, as well as recovery from, selected stress events (negative pointer years). In addition, I studied δ13C signatures of tree rings as proxy for stomatal conductance..The seed bank of stands with moderately diverse and highly diverse tree layer is not only richer in species than the monospecific stands, but also in the number of germinable seeds. In spite of the strong correlation found between tree- and herb-layer diversity in the stands, an indirect effect of tree-layer diversity or tree species identity on the seed bank via herb-layer composition could not be confirmed due to the poor correspondence between herb-layer composition and seed bank structure. The effect exerted by certain tree species on litter quality and decomposition rates had a stronger influence on the seed bank, which is indicated by the strong negative correlation between seed bank size and litter layer thickness.The impact of neighborhood diversity was also visible in Fagus trees showing higher mean radial stem growth rates when they were surrounded by a number of other tree species. My study is among the first that obtained evidence for a species-specific neighborhood effect on the current wood increment rate of a target species in natural forests. Tree-species identity and the contribution of different species to the competition index (CI) of a Fagus neighborhood were more important than the magnitude of CI itself. Beeches with a few allospecific neighbors (Fagus70-99 category) showed less severe growth reduction in the negative pointer year 1976 than beeches in pure stands. They also recovered more rapidly in terms of stem increment after the severe 1976 drought. The buffering effect of allospecific neighborhoods with regard to environmental hazards could explain this finding.In the period from 1926 to 1975, beech predominantly surrounded by conspecifics (Fagus100, Fagus70-99) reached higher values of δ13C than beech largely affected by allospecific competitors (Fagus<70). On the contrary, values of the more actual period 1976 to 2005 did not vary between the neighborhood classes. This leads to the conclusion that target trees, today surrounded by allospecific neighbors, probably benefited from thinning in order to promote growth of other valuable broad-leaved tree species. Thinning resulted in an increase in available soil water and growth. This increase in soil water is manifested in relatively smaller δ13C values in the largely allospecific Fagus<70 neighborhoods. The effect vanished when the canopy closed again. Hence, the observed δ13C pattern is probably rather a secondary effect of forest management than a direct effect of neighborhood diversity on water supply of Fagus.Disturbance regime, forest management and stand age are important for ecosystem processes in forests. They leave detectable traces in the soil seed bank, tree-ring width and δ13C series. Generally, I assume that the neighborhood represents an environmental factor of lower importance than other factors controlling annual ring width, as there are soil conditions and light availability. Thus, the neighborhood may influence the growth response under certain conditions, but this factor is too weak to determine principal types in the growth-environment relationship of Fagus.