Growth Trends Driven By Environmental Factors Extracted From Long Term Experimental Data In Southwest Germany

Recently, two new methods have been developed to extract environmentally driven growth trends from data provided by long-term experiments. The methods draw either on year ring series of sample trees cut during the experiments (annually resolved growth data, tree-level diameter increment trend; (YuE et al., 2011) or on series of periodic re-measurements of the whole stands (multi-annual measurement intervals, stand-level basal area increment trends; (YuE et al., 2012). Data from Norway spruce, silver fir, Douglas-fir or European beech for exemplifying purposes demonstrate that both methods result in comparable trends (figure 5). Based on these findings, we exploited in this study the rather substantial database available for a variety of tree species from periodically repeated stand measurements of long-term experimental sites in southwest Germany (cf. table 1, figures 1-2) for the purposes of growth trend analyses.Overall, the following general pattern of growth trends in southwest Germany emerged: all tree species included in the analysis (Norway spruce, silver fir, Douglas-fir, Scots pine, European beech, and oak) showed increment trends increasing from the middle of the 20th century until the 1990s (cf. figures 5-6). For silver fir (figure 5), this increasing trend was temporarily disrupted by a characteristic growth depression occurring in the 1970s-1980s. In the 1990s, growth trends reversed in all tree species into declining trends. For comparative purposes, trends for temperature and precipitation were averaged from all experiment locations. However, both climate factors (cf. figure 3, left & central) did not correspond with growth trends. In particular, they were lacking explanatory power for the general trend reversion observed in the 1990s. Therefore, the DeMartonne index for aridity was additionally calculated as an attempt to capture the influence of the interaction of these two climatic factors on water supply: In contrast to the temperature or precipitation trends, the trend of the index of aridity corresponded not only with increasing growth trends since the 1950s but also with the reversion into declining trends in the 1990s (cf. figure 3, right). However, the magnitude of temporal variation in growth trends (cf. figures 5-6) clearly exceeded that of the index of aridity, which was only marginal if calculated for the vegetation period (cf. figure 3, right). Possibly, this indicates that in addition to climate factors other growth relevant environmental factors have impacted on the growth trends.In order to investigate possible regional differences in growth and climate trends, experiment locations were separated according to their respective temperature regime indicating their affiliation to climatic elevation zones (plains of low altitude high mountainous zones) (WOLFF et al., 2003). Obviously, the increase in temperature trend has relatively been strongest in the cooler regions (cf. figure 4, left), whereas the trend decline since the 1990s of the index of aridity has not displayed much differentiation between the regions (cf. figure 4, right). Interestingly growth trends differentiated for regions indicate a relationship between climate characteristics in a tree species' major natural range and current development in growth trends. Tree species naturally occurring in relatively cooler climate (conifers: Norway spruce and Scots pine; deciduous: European beech) current growth trends observed in relatively cooler regions of southwest Germany exceed the development in relatively warmer regions (cf. figure 7). In contrast, the growth trend of oak in the warmer region exceeded its trend in the cooler zone (cf. figure 7). Silver fir shows a more intermediate position as growth trends do not differ much between the warmer and the cooler zone (cf. figure 8).As the magnitude of the temporal variation of growth trends clearly exceeds the magnitude of the differences observed between the regions we conclude that it will be particularly important to address the aspect of temporal variation of growth and growth relevant factors adequately in causal investigations as well as in developing environment sensitive growth models.