Reams and Huso (1990) also noted that declining red spruce stands in Maine
were released one to three decades before dieback started. Historical logging may
thus have caused sudden changes in the growth dynamics of surviving trees leading
to drought-induced hydraulic failure. Also, microclimatic conditions (e.g. air and
soil humidity) change drastically in logged open stands as compared with closed
forests which may also affect silver-fir performance (Aussenac 2002). For instance,
silver-fir defoliation increases in habitats with low soil-water holding capacity
(Thomas et al. 2002). The 1980s releases found only in the declining sites were
caused by the felling of dying trees, but many of the surviving trees did not improve
their growth in the long term and, in many cases, died. These facts suggest that their
performance was permanently affected by drought stress leading to an irreversible
reduction in growth and suggesting a loss in stomatal regulation of declining trees
despite the recent rise in atmospheric CO 2 concentrations (Linares and Camarero
2012). Additional research is required to establish the links between soil-water
availability, growth trends and dieback if these processes are to be used as monitors
of the effects of climate warming on mountain forests.
6.5 Climate-Growth Associations: The Critical Role
Played by Late-Summer Water Deficit
Silver-fir growth was enhanced by wet spring conditions and by cool and wet
conditions during the summer prior to tree-ring formation (Fig. 6.7a). The first two
axes of the PCA based on the correlation coefficients between climatic variables and
site chronologies accounted for 26.7% and 17.0% of the total variance, respectively,
(Fig. 6.7b). These analyses detected a positive and stronger response of tree growth
in declining than in non-declining sites to current June precipitation, and a lower
response to previous September precipitation and February temperature (negative
relationship). The positive effect of current June (previous September) precipitation
on growth in declining sites was significantly stronger (lower) than in non-declining
sites (June, F = 6.57; September, F = 6.30, P < 0.02 in both cases). The cumulative water deficits of the previous spring and the current growing season (January–
May) were also negatively related to growth in declining sites, and these effects
were more marked westwards. Silver-fir growth was also negatively (positively)
associated with higher September (April) temperatures of the previous (current)
year, and this association increased eastwards (westwards).
Climate-growth relationships were unstable through time since most
growth-climate relationships changed in the 1980s according to moving correlations functions (Fig. 6.8). The negative influence of previous February and
September temperatures and the positive influence of previous September precipitation on silver-fir growth strengthened since 1985 in the non-declining sites,
whereas the positive influence of current July precipitation decreased. Declining
sites showed a similar temporal instability of their growth-climate relationships,
146
J.J. Camarero
were released one to three decades before dieback started. Historical logging may
thus have caused sudden changes in the growth dynamics of surviving trees leading
to drought-induced hydraulic failure. Also, microclimatic conditions (e.g. air and
soil humidity) change drastically in logged open stands as compared with closed
forests which may also affect silver-fir performance (Aussenac 2002). For instance,
silver-fir defoliation increases in habitats with low soil-water holding capacity
(Thomas et al. 2002). The 1980s releases found only in the declining sites were
caused by the felling of dying trees, but many of the surviving trees did not improve
their growth in the long term and, in many cases, died. These facts suggest that their
performance was permanently affected by drought stress leading to an irreversible
reduction in growth and suggesting a loss in stomatal regulation of declining trees
despite the recent rise in atmospheric CO 2 concentrations (Linares and Camarero
2012). Additional research is required to establish the links between soil-water
availability, growth trends and dieback if these processes are to be used as monitors
of the effects of climate warming on mountain forests.
6.5 Climate-Growth Associations: The Critical Role
Played by Late-Summer Water Deficit
Silver-fir growth was enhanced by wet spring conditions and by cool and wet
conditions during the summer prior to tree-ring formation (Fig. 6.7a). The first two
axes of the PCA based on the correlation coefficients between climatic variables and
site chronologies accounted for 26.7% and 17.0% of the total variance, respectively,
(Fig. 6.7b). These analyses detected a positive and stronger response of tree growth
in declining than in non-declining sites to current June precipitation, and a lower
response to previous September precipitation and February temperature (negative
relationship). The positive effect of current June (previous September) precipitation
on growth in declining sites was significantly stronger (lower) than in non-declining
sites (June, F = 6.57; September, F = 6.30, P < 0.02 in both cases). The cumulative water deficits of the previous spring and the current growing season (January–
May) were also negatively related to growth in declining sites, and these effects
were more marked westwards. Silver-fir growth was also negatively (positively)
associated with higher September (April) temperatures of the previous (current)
year, and this association increased eastwards (westwards).
Climate-growth relationships were unstable through time since most
growth-climate relationships changed in the 1980s according to moving correlations functions (Fig. 6.8). The negative influence of previous February and
September temperatures and the positive influence of previous September precipitation on silver-fir growth strengthened since 1985 in the non-declining sites,
whereas the positive influence of current July precipitation decreased. Declining
sites showed a similar temporal instability of their growth-climate relationships,
146
J.J. Camarero