Water and Nutrient Limitations to Primary Production
191
13.6 Follow-up to the Experiment
Our experiment was designed to run for 3 years, and irrigation was discontinued after the third summer (1991). However, in March 1992 fertilizers
were applied again in the same plots and doses as in 1989, in order to maintain a high nutrient availability for other studies. The fourth year after the
first treatments (1992) happened to be one of exceptionally high springsummer rainfall. We took advantage of this opportunity to assess the tree
growth and litterfall responses to a natural reduction in summer water stress,
and to test whether such responses were affected by irrigation during the
previous 3 years and by fertilization. We were particularly interested in comparing the effects of high rainfall during the warm season to those of experimental irrigation, and testing whether a high light-interception capacity
(as induced by previous irrigation and by N fertilization) had a carry-on
positive effect on stem diameter growth. So, stem diameters of all trees were
re-measured at the end of this growing season (December 1992), and
monthly litterfall collections were continued through 1992.
In control plots, mean stem diameter growth of holm oak during 1992 was
0.67 mm year- I (± 0.19 SE among plot means), a growth rate doubling that of
the 3 preceding years (Mayor et al. 1994). Diameter growth in control plots in
1992 was higher than that achieved during 1989-1991 in the irrigated plots
(mean of the I and IP treatments: 0.54 mm year-I). These two observations
clearly indicate that 1992 was a much better year for holm oak growth than
the 3 previous years, and that soil water availability during the growing season was probably increased more by the abundant spring-summer rainfall of
1992 than it had been by irrigation in the preceding years. Annual rainfall in
1992 was 855 mm, while mean rainfall plus irrigation in 1989-1991
amounted to only 735 mm. The inability of forest irrigation to equal the effect of natural precipitation has been also observed in experiments in other
forest types (Lucier and Hinckley 1982; Gower et al. 1992).
When all holm oak stems (both dominant and suppressed) were included
in the analysis, none of the three experimental factors (previous irrigation, N
fertilization, and P fertilization) significantly affected stem diameter growth
during 1992. For dominant stems, N fertilization significantly increased mean
stem diameter growth in 1992 by 42%, compared to non-fertilized stems,
while P fertilization and previous irrigation did not. Thus, increased light
interception capacity by itself (as shown by the previously irrigated plots)
was not conducive to increased stem diameter growth in a year of relatively
high water availability. N fertilization increased the diameter growth of
dominant trees in 1992, an effect which had not taken place during the 3 previous years. Such a different response could arise from (1) fertilizer being
applied a second time in March 1992, (2) a time-delayed response to the first
fertilization, or (3) the effects of improved N nutrition or increased light interception capacity coupled with a high soil water availability during the
growth season of 1992.
191
13.6 Follow-up to the Experiment
Our experiment was designed to run for 3 years, and irrigation was discontinued after the third summer (1991). However, in March 1992 fertilizers
were applied again in the same plots and doses as in 1989, in order to maintain a high nutrient availability for other studies. The fourth year after the
first treatments (1992) happened to be one of exceptionally high springsummer rainfall. We took advantage of this opportunity to assess the tree
growth and litterfall responses to a natural reduction in summer water stress,
and to test whether such responses were affected by irrigation during the
previous 3 years and by fertilization. We were particularly interested in comparing the effects of high rainfall during the warm season to those of experimental irrigation, and testing whether a high light-interception capacity
(as induced by previous irrigation and by N fertilization) had a carry-on
positive effect on stem diameter growth. So, stem diameters of all trees were
re-measured at the end of this growing season (December 1992), and
monthly litterfall collections were continued through 1992.
In control plots, mean stem diameter growth of holm oak during 1992 was
0.67 mm year- I (± 0.19 SE among plot means), a growth rate doubling that of
the 3 preceding years (Mayor et al. 1994). Diameter growth in control plots in
1992 was higher than that achieved during 1989-1991 in the irrigated plots
(mean of the I and IP treatments: 0.54 mm year-I). These two observations
clearly indicate that 1992 was a much better year for holm oak growth than
the 3 previous years, and that soil water availability during the growing season was probably increased more by the abundant spring-summer rainfall of
1992 than it had been by irrigation in the preceding years. Annual rainfall in
1992 was 855 mm, while mean rainfall plus irrigation in 1989-1991
amounted to only 735 mm. The inability of forest irrigation to equal the effect of natural precipitation has been also observed in experiments in other
forest types (Lucier and Hinckley 1982; Gower et al. 1992).
When all holm oak stems (both dominant and suppressed) were included
in the analysis, none of the three experimental factors (previous irrigation, N
fertilization, and P fertilization) significantly affected stem diameter growth
during 1992. For dominant stems, N fertilization significantly increased mean
stem diameter growth in 1992 by 42%, compared to non-fertilized stems,
while P fertilization and previous irrigation did not. Thus, increased light
interception capacity by itself (as shown by the previously irrigated plots)
was not conducive to increased stem diameter growth in a year of relatively
high water availability. N fertilization increased the diameter growth of
dominant trees in 1992, an effect which had not taken place during the 3 previous years. Such a different response could arise from (1) fertilizer being
applied a second time in March 1992, (2) a time-delayed response to the first
fertilization, or (3) the effects of improved N nutrition or increased light interception capacity coupled with a high soil water availability during the
growth season of 1992.