Bilan de santé actuel et rétrospectif du sapin (Abies alba Mill.) dans les Vosges. Etude écologique et dendrochronologique

Abstract

One hundred and ninety six stands were chosen within the whole natural range of white fir in the region, in order to include all the site types according to age, elevation, soil, topography, vegetation, but without prejudice to the apparent health status of the trees. Six dominant and codominant trees were selected in each stand. Their foliage density was noted and a core was taken to the pith at breast height. Thus, almost 1,200 cores, i.e. about 120,000 annual growth rings, were used for the dendrochronological study. The analysis was made difficult because of the frequent occurence of "missing rings", which must be detected in order to avoid desynchronization of the whole previous succession of rings. This was done by crossdating, using characteristic years (as far as the ring thickness is concerned) as reference years: 1870, 1893, 1916, 1922, 1934, 1948, 1956, 1961, 1976... We tried to test the hypothesis that an increasing level of atmospheric pollution, from a date to be determined, would be the cause of a decrease in growth, which might have been developing into a real dieback for some years. From the many data available, all the rings produced at a current reference age were considered. Because of the wide diversity of the present age of the stands, the rings correspond to various calendar years. Whatever the reference age (60 to 110 years), no significant general trend in radial growth variation could be noticed since the beginning of the 20th century. As soon as this stage of interpretation, it was possible to assert that the fir forest in the Vosges was not in danger of dying out before long. This was contrary to the strong fear which prevailed before this study. However, this did not preclude that this average reality could hide more alarming special situations susceptible to spread more widely in a near future. In order to assess this possibility, the initial sample of plots was subdivided into several subsamples according to some ecological factors : elevation, soil fertility, exposure, topography. All the radial growth data were previously converted into growth indices avoiding the influence of the current age. This transformation was done in reference to the average curve "ring width/age" of the whole sample. A surprising but irrefutable conclusion emerges: none of the parameters which were used makes possible to distinguish tree populations with a behaviour significantly different from the mean. However, all the curves show that the fir forests in the Vosges have actually suffered a period of severe stress which began as early as 1973, culminated in 1976, but was almost finished in 1983 (at least, as far as radial growth is concerned). The same curves also show previous crises, comparable to the last one for their extension and severity: 1943-1951, centered on 1948, and 1917-1925, centered on 1922. As it is the case with 1976, these periods correspond to years or succession of years with a pronounced rainfall deficit. Furthermore, the aspect of crowns and the estimated loss of needles are poor indicators of the actual vigor of the trees: the growth curves significantly deviate from the average only for needle losses above 40 p. 100. On the contrary, sapwood width is a much more consistent parameter : it is closely correlated with the present growth, and the radial growth curves of trees in classes with different sapwood width have been deviating for 50 to 60 years, practically since the 1917-1925 crisis The impact of an increasing large-scale air pollution (possible but not proved) cannot be left out, but the cause of the present crisis proves to be mainly climatic, as for the previous ones. Individual differences in tree response seem to be mostly due to the structure of the stands.