In both years the time-course of the total dry matter
accumulation was significantly affected by the interaction
treatment×time. That is, despite water deficit
was 318Lm−2 (1999) and 324 Lm−2 (2000) higher
in W0 (Fig. 1), the growth rate of W1 was faster only
in the early growth period. The course of growth was
significantly affected both by physiological and morphological
modifications. Furthermore, morphological
components were found to be mostly influenced by the
specific leaf area (SLA, i.e. the total leaf area per unit
leaf mass). Specifically, until 1200GDDSLA was 20%
(year 1999) and 16% (2000) higher in W1 than in W0.
Water regime also significantly affected the biomass
partitioning (Fig. 2). In both years, the highest leaf dry
weights (LDW, i.e. the total leaf mass per unit total
plant mass) occurred at about 1300GDD parallel with
the leaf area index (LAI). LDW and LAI were slightly
higher in the first year, however they were much more
affected by treatment than by year (Fig. 2). The differences
between treatments on LDW, as well as on LAI,
were mostly explained by the two first irrigations. Differences
were however much less marked on LDW than
on LAI. Thus, leaf extension more than leaf biomass
was affected by the treatment according to SLA variations.
Water regime generally favoured stem growth
(SDW) as well. During summertime the average SDW
of W1 were 34% (1999) and 45% (2000) higher than
that ofW0 (Fig. 2). A different trend between years was
also observed on leaf and stem developments: SDW was generally higher in the second year while LDW
was higher in 1999. The tuber dry weight (TDW) was
lower inW1 over the major part of the crop cycle, while
in the last part of the growing season, a very fast TDW
occurred inW1. This allowedW1 to reach a significantly
higher (1999) or similar (2000) TDWthanW0 (Fig. 2).
Differences in TDW between treatments mainly concerned
the number of tubers while tuber sizewas poorly
affected (data not reported).
Since RUE was not significantly affected by year,
the data of 2 years were grouped (Fig. 3). RUE was
significantly influenced by treatment, W1 showing a
higher value than W0. If RUE is calculated on fructan
accumulation (RUEf), it may indicate the light use efficiency
on end-products (Fig. 3). An opposite trend of
RUEf was observed between W1 and W0. That is, until
900 MJm−2, RUEf of W0 was about three-fold the
irrigated plants; afterward RUEf of W0 was only 40%
of W1