64
capacity for adaptation to eventual environmental changes, and for selective breeding. However, each population should be considered separately for each trait.
On the other hand, differences between populations were significant for all variables except for the ratio between root and stem dry biomasses. Quantitative differentiation between populations was moderate on average (Q STmean = 36%) and
higher than neutral differentiation (F ST between 5 and 6%, Marchelli and Gallo
2001; Azpilicueta et al. 2013), thus indicating the likely occurrence of an adaptation
process to current environmental conditions. According to post hoc comparisons,
the northernmost of the analysed populations (Tromen) was the most different; a
result that is consistent with previous studies based on chloroplast DNA markers
(Marchelli et al. 1998; Marchelli and Gallo 2006), on SSRs (Azpilicueta et al. 2013)
and on seed traits (Marchelli and Gallo 1999).
Still thinking on juvenile characters, a high-density field trial (Trial 2, Table 3.1)
was installed in 2011 in San Martín Forest Station of INTA (Instituto Nacional de
Tecnología Agropecuaria) for early evaluation. It consisted of 1290 2-year-old seedlings from 86 open-pollinated families, corresponding to 8 Argentinian natural populations, arranged in a randomized complete block design (Table 3.1). Four and five
years later, during the growing seasons, the total height of each sapling was measured repeatedly once a week the first year and every other week the second year
(Duboscq-Carra 2019). The genetic variation for daily growth rate (mean of the
whole trial was 2.4 mm/d and 3.1 mm/d for the first and the second seasons, respectively) was analysed by means of mixed linear models. The effect of “population”
and “family” factors were not significant in any of the 2 years. Subsequently, individual growth curves were regressed (Boltzmann model) for each sapling, and the
genetic variation of some essential parameters of the adjusted curves was analysed.
On average, growth initiated (t 10 : time to reach 10% of the season growth) in the
278 day of the year (DOY) (October 4) and in the 301 DOY (October 27) in the
fourth and the fifth seasons, respectively. On the other hand, growth ceased (t 90 : time
to reach 90% of the season growth) in the 353 DOY (December 18) and 363 DOY
(December 27) in each season, respectively. Differences were observed among populations for t 10 in both seasons, but only in the second season for t 90 . Differentiation
was estimated as moderate for t 10 in the first (Q ST = 15%) and in the second
(Q ST = 17%) seasons but low for t 90 in the second season (Q ST = 9%; Table 3.1). The
family factor was not significant in any of the seasons for any of these two traits.
Among the adaptive traits, the phenological ones are probably the most affected
by global climate change (Bertin 2008). In trees of temperate zones, phenology of
processes such as bud burst and bud set, or leaf senescence, responds to a balance
between survival and productivity. Since the risk of frost damage (necrosis by freezing of primary meristems) is more frequent in autumn, trees adapted to colder climates complete their growth early in the season, forming precociously the winter
resistance structures (buds) (Howe et al. 2003). The cessation of growth, the formation of buds and the subsequent senescence of leaves occur mainly by changes in the
photoperiod (longer nights) and lower temperatures (colder nights). On the other
hand, the opening of buds, once the increasing photoperiod signal has occurred
(about June 22 in the Southern Hemisphere), is largely influenced by the
P. Marchelli et al.
capacity for adaptation to eventual environmental changes, and for selective breeding. However, each population should be considered separately for each trait.
On the other hand, differences between populations were significant for all variables except for the ratio between root and stem dry biomasses. Quantitative differentiation between populations was moderate on average (Q STmean = 36%) and
higher than neutral differentiation (F ST between 5 and 6%, Marchelli and Gallo
2001; Azpilicueta et al. 2013), thus indicating the likely occurrence of an adaptation
process to current environmental conditions. According to post hoc comparisons,
the northernmost of the analysed populations (Tromen) was the most different; a
result that is consistent with previous studies based on chloroplast DNA markers
(Marchelli et al. 1998; Marchelli and Gallo 2006), on SSRs (Azpilicueta et al. 2013)
and on seed traits (Marchelli and Gallo 1999).
Still thinking on juvenile characters, a high-density field trial (Trial 2, Table 3.1)
was installed in 2011 in San Martín Forest Station of INTA (Instituto Nacional de
Tecnología Agropecuaria) for early evaluation. It consisted of 1290 2-year-old seedlings from 86 open-pollinated families, corresponding to 8 Argentinian natural populations, arranged in a randomized complete block design (Table 3.1). Four and five
years later, during the growing seasons, the total height of each sapling was measured repeatedly once a week the first year and every other week the second year
(Duboscq-Carra 2019). The genetic variation for daily growth rate (mean of the
whole trial was 2.4 mm/d and 3.1 mm/d for the first and the second seasons, respectively) was analysed by means of mixed linear models. The effect of “population”
and “family” factors were not significant in any of the 2 years. Subsequently, individual growth curves were regressed (Boltzmann model) for each sapling, and the
genetic variation of some essential parameters of the adjusted curves was analysed.
On average, growth initiated (t 10 : time to reach 10% of the season growth) in the
278 day of the year (DOY) (October 4) and in the 301 DOY (October 27) in the
fourth and the fifth seasons, respectively. On the other hand, growth ceased (t 90 : time
to reach 90% of the season growth) in the 353 DOY (December 18) and 363 DOY
(December 27) in each season, respectively. Differences were observed among populations for t 10 in both seasons, but only in the second season for t 90 . Differentiation
was estimated as moderate for t 10 in the first (Q ST = 15%) and in the second
(Q ST = 17%) seasons but low for t 90 in the second season (Q ST = 9%; Table 3.1). The
family factor was not significant in any of the seasons for any of these two traits.
Among the adaptive traits, the phenological ones are probably the most affected
by global climate change (Bertin 2008). In trees of temperate zones, phenology of
processes such as bud burst and bud set, or leaf senescence, responds to a balance
between survival and productivity. Since the risk of frost damage (necrosis by freezing of primary meristems) is more frequent in autumn, trees adapted to colder climates complete their growth early in the season, forming precociously the winter
resistance structures (buds) (Howe et al. 2003). The cessation of growth, the formation of buds and the subsequent senescence of leaves occur mainly by changes in the
photoperiod (longer nights) and lower temperatures (colder nights). On the other
hand, the opening of buds, once the increasing photoperiod signal has occurred
(about June 22 in the Southern Hemisphere), is largely influenced by the
P. Marchelli et al.