Rolland et al. 1999; Tardif et al. 2003). This fact is consistent with the low stomatal
regulation of gas exchange in silver fir and its dieback in sites with high leaf-to-air
vapour pressure difference in response to warmer air temperatures (Peguero–Pina
et al. 2007; Vicente–Serrano et al. 2015). Other factors such as nitrogen deposition
might also be involved as occurred in the Vosges, where silver-fir dieback was also
related to acidification (Pinto et al. 2007). However, nitrogen deposition is much
higher there than in the Aragón Pyrenees (de Vries et al. 2003), and most of the
study sites were located on basic soils and showed N-NW exposure.
The negative response of silver-fir growth to the growing-season cumulative
water deficit increased westwards and was more marked in declining than in
non-declining sites. Radial growth was also enhanced by current June precipitation
and this effect also increased westwards as summer precipitation decreased
(Figs. 6.3 and 6.7). Accordingly, growth in the western declining sites was more
constrained by climatic conditions during the year of tree-ring formation than in the
eastern non-declining sites. However, growth in declining sites was also affected by
distinctive factors such as the negative relationship with previous February
temperature.
Contrastingly, an increasing response of growth to previous September climatic
conditions eastwards was detected, whereas the negative (positive) relationship with
temperature (precipitation) increased during the second half of the past century
(Fig. 6.8). These results seem to be counter intuitive since the 1986 dieback onset in
the western study area was preceded by an extremely dry and warm September.
Indeed, 1985 was the year with lowest September precipitation in the western study
area during the twentieth century, and such dryness was caused by the presence of
two high pressures over the northern Atlantic.
In the second half of the past century, silver-fir growth was positively and
significantly related to the Scandinavian (r = 0.41) and the North Atlantic
Oscillation (r = 22) indices of the previous September and November, respectively.
The associations with these two dominant modes of atmospheric variability in SW
Europe (Sáenz et al. 2001) were unstable since the correlations with the
Scandinavian index peaked in the 1967–1986 period, whereas the NAO effects have
changed from positive to negative. This transition was clearer for the western
declining sites than elsewhere which suggests some connection between atmospheric variability and silver-fir dieback (Camarero 2011). The lower responsiveness to previous September climate in the westernmost area was linked to an earlier
worsening of late-summer conditions there, probably through an intensification of
warming-induced drought stress.
The presented findings have implications for silver-fir growth and persistence
near the rear xeric edge of the species’ distribution area. First, the year-to-year
variation in tree growth of silver fir has increased in the second half of the twentieth
century in response to a greater warming-induced drought stress and the occurrence
of more frequent dry spells in the last decades as compared with earlier more
favourable conditions in the first half of the past century (Tardif et al. 2003). It is
also evidenced that this instability also affected the associations between growth
and atmospheric variability since 1950. Second, it is expected that, in some
6 The Multiple Factors Explaining Decline in Mountain Forests …
149
regulation of gas exchange in silver fir and its dieback in sites with high leaf-to-air
vapour pressure difference in response to warmer air temperatures (Peguero–Pina
et al. 2007; Vicente–Serrano et al. 2015). Other factors such as nitrogen deposition
might also be involved as occurred in the Vosges, where silver-fir dieback was also
related to acidification (Pinto et al. 2007). However, nitrogen deposition is much
higher there than in the Aragón Pyrenees (de Vries et al. 2003), and most of the
study sites were located on basic soils and showed N-NW exposure.
The negative response of silver-fir growth to the growing-season cumulative
water deficit increased westwards and was more marked in declining than in
non-declining sites. Radial growth was also enhanced by current June precipitation
and this effect also increased westwards as summer precipitation decreased
(Figs. 6.3 and 6.7). Accordingly, growth in the western declining sites was more
constrained by climatic conditions during the year of tree-ring formation than in the
eastern non-declining sites. However, growth in declining sites was also affected by
distinctive factors such as the negative relationship with previous February
temperature.
Contrastingly, an increasing response of growth to previous September climatic
conditions eastwards was detected, whereas the negative (positive) relationship with
temperature (precipitation) increased during the second half of the past century
(Fig. 6.8). These results seem to be counter intuitive since the 1986 dieback onset in
the western study area was preceded by an extremely dry and warm September.
Indeed, 1985 was the year with lowest September precipitation in the western study
area during the twentieth century, and such dryness was caused by the presence of
two high pressures over the northern Atlantic.
In the second half of the past century, silver-fir growth was positively and
significantly related to the Scandinavian (r = 0.41) and the North Atlantic
Oscillation (r = 22) indices of the previous September and November, respectively.
The associations with these two dominant modes of atmospheric variability in SW
Europe (Sáenz et al. 2001) were unstable since the correlations with the
Scandinavian index peaked in the 1967–1986 period, whereas the NAO effects have
changed from positive to negative. This transition was clearer for the western
declining sites than elsewhere which suggests some connection between atmospheric variability and silver-fir dieback (Camarero 2011). The lower responsiveness to previous September climate in the westernmost area was linked to an earlier
worsening of late-summer conditions there, probably through an intensification of
warming-induced drought stress.
The presented findings have implications for silver-fir growth and persistence
near the rear xeric edge of the species’ distribution area. First, the year-to-year
variation in tree growth of silver fir has increased in the second half of the twentieth
century in response to a greater warming-induced drought stress and the occurrence
of more frequent dry spells in the last decades as compared with earlier more
favourable conditions in the first half of the past century (Tardif et al. 2003). It is
also evidenced that this instability also affected the associations between growth
and atmospheric variability since 1950. Second, it is expected that, in some
6 The Multiple Factors Explaining Decline in Mountain Forests …
149