Finally, to summarise these results and to define the main climatic response of
declining versus non-declining sites, I performed a PCA on the matrix of the
correlation function coefficients.
6.2 Warmer Climate Conditions and the 1980s Peak
in Water Deficit
Overall, the climate is becoming warmer in the study area with a peak in drought
stress during the 1980s. A strong warming trend was observed in the WAP during
the late 1940s and in the 1980s (Fig. 6.2a). Precipitation decreased in the 1980s
and increased during the 1960s. Furthermore, the relative contribution of summer
rainfall to the total annual precipitation has significantly declined in the WAP area
during the last half of the twentieth century (trend = −0.09; P < 0.01). The mean
annual water deficit during the 1940–1999 period was significantly higher in the
WAP (82.3 ± 8.3 mm) than in the EAP (49.6 ± 7.0 mm) sub-region. The estimated annual water deficit peaked in 1985 and 1967 in both sub-regions but
always reaching a greater annual deficit in the WAP sub-region than eastwards
(Fig. 6.2b).
6.3 Structural Features of Silver-Fir Stands Presenting
Dieback: Low Growth Rates
Most sites showing severe defoliation and dieback were located in the WAP
sub-region at mid elevation (Table 6.1; Fig. 6.3b). These stands were dominated by
silver-fir trees with low dbh and ages ranging 90–155 years old and corresponded
to mixed forests with a low basal area. On average, we found a significantly higher
frequency (Mann–Whitney test, U = 7.5, P < 0.001) of dead trees in declining
(10.8 ± 1.1%) than in non-declining sites (1.9 ± 0.4%; Table 6.1).
The low productivity of declining sites is confirmed by the fact that radial
growth as a function of cambial age was consistently lower there than in
non-declining sites (Fig. 6.4). The most defoliated trees from declining sites also
tended to show less sapwood area (r s = −0.26, P < 0.05). Declining sites were also
those where more releases were detected since we found a significant positive
relationship (r s = 0.88, P < 0.001) between the stand mean defoliation and the
average number of releases per tree (Fig. 6.5a). For instance, 58% of defoliated
(crown defoliation > 50%) trees from declining sites showed at least two releases
before the 1980s, whereas only 35% of trees without defoliation from non-declining
sites did (Fig. 6.5b).
140
J.J. Camarero
declining versus non-declining sites, I performed a PCA on the matrix of the
correlation function coefficients.
6.2 Warmer Climate Conditions and the 1980s Peak
in Water Deficit
Overall, the climate is becoming warmer in the study area with a peak in drought
stress during the 1980s. A strong warming trend was observed in the WAP during
the late 1940s and in the 1980s (Fig. 6.2a). Precipitation decreased in the 1980s
and increased during the 1960s. Furthermore, the relative contribution of summer
rainfall to the total annual precipitation has significantly declined in the WAP area
during the last half of the twentieth century (trend = −0.09; P < 0.01). The mean
annual water deficit during the 1940–1999 period was significantly higher in the
WAP (82.3 ± 8.3 mm) than in the EAP (49.6 ± 7.0 mm) sub-region. The estimated annual water deficit peaked in 1985 and 1967 in both sub-regions but
always reaching a greater annual deficit in the WAP sub-region than eastwards
(Fig. 6.2b).
6.3 Structural Features of Silver-Fir Stands Presenting
Dieback: Low Growth Rates
Most sites showing severe defoliation and dieback were located in the WAP
sub-region at mid elevation (Table 6.1; Fig. 6.3b). These stands were dominated by
silver-fir trees with low dbh and ages ranging 90–155 years old and corresponded
to mixed forests with a low basal area. On average, we found a significantly higher
frequency (Mann–Whitney test, U = 7.5, P < 0.001) of dead trees in declining
(10.8 ± 1.1%) than in non-declining sites (1.9 ± 0.4%; Table 6.1).
The low productivity of declining sites is confirmed by the fact that radial
growth as a function of cambial age was consistently lower there than in
non-declining sites (Fig. 6.4). The most defoliated trees from declining sites also
tended to show less sapwood area (r s = −0.26, P < 0.05). Declining sites were also
those where more releases were detected since we found a significant positive
relationship (r s = 0.88, P < 0.001) between the stand mean defoliation and the
average number of releases per tree (Fig. 6.5a). For instance, 58% of defoliated
(crown defoliation > 50%) trees from declining sites showed at least two releases
before the 1980s, whereas only 35% of trees without defoliation from non-declining
sites did (Fig. 6.5b).
140
J.J. Camarero