376
J.R. Ehleringer
hydraulic conductances than annuals. Long-lived perennials, which must
persist through repeated droughts of varying duration, would be expected
to have lower hydraulic conductances than short-lived perennials in order
to avoid extensive xylem cavitation. With respect to perennial life forms
accessing the same limited soil moisture, trees would be at a significant
disadvantage to shrubs (an overall shorter conducting system). On this
point Balding and Cunningham (1974) showed that plant height became
progressively shorter in Acacia constricta (tree) and Larrea tridentata (shrub)
as drought severity increased along an aridity gradient. Stem xylem cavitation
events would be expected to proceed from the stem tip where transpiration
is occurring toward basal regions, creating progressive stem dieback under
conditions of increasing drought duration. This may in part be the explanation for why suffrutescent growth (many stems emerging from a common
root base) is so common among aridland shrubs.
Life-form diversity of perennial species is positively related to climatic
diversity. Cody (1989) showed that life-form diversity among Sonoran Desert
sites increased under extreme conditions - conditions likely to reflect both
extremes in drought and a high year-to-year variability in precipitation.
Shmida and Burgess (1988) found that biseasonality of precipitation was the
major driving factor accounting for the increased growth-form diversity
in the Sonoran Desert. Biseasonal precipitation was also correlated with
these same life-form patterns within the Great Basin and Colorado Plateau
(Comstock and Ehleringer 1992). Together, these data suggest that in North
American deserts it is the diversity of incoming moisture sources that drives
the evolution of differences in life-form.
The coefficient of variation is a measure of unpredictability (Sect. 18.2.1)
and an unpredictable environment will preclude the evolution of perennial
plants dependent on frequent rainfall episodes or not capable of persisting
through extended drought periods. The loss of constancy (the degree to
which conditions are similar from year to year) in drier sites should have a
direct impact on the distribution of life-forms along an aridity gradient.
Hastings and Turner (1965a) examined the amounts and variability of winter
and summer precipitation in Baja California. They noted that the variability
in precipitation corresponded to the distribution of vegetation types described
by Shreve (1934) and Shreve and Wiggins (1964). After Colwell (1974)
showed that predictability and constancy were mathematically related,
Schaffer and Gadgil (1975) then followed with a prediction of the distribution
abundances of annual versus perennial life-forms as predictability of rainfall
decreased. The data of Harper (unpub!' data cited in Schaffer and Gadgil
1975) supported the prediction that annuals become an increasing fraction of
the total flora as predictability of that precipitation decreases. What these
data sets did not reveal was any information on the distribution of photosynthetic pathways among these annuals. Such information can be extracted
from Shreve and Wiggins (1964), who noted that there were distinct winterand summer-annual floras, even within a particular region. It turns out that
winter annuals have C 3 photosynthesis exclusively, while summer annuals
J.R. Ehleringer
hydraulic conductances than annuals. Long-lived perennials, which must
persist through repeated droughts of varying duration, would be expected
to have lower hydraulic conductances than short-lived perennials in order
to avoid extensive xylem cavitation. With respect to perennial life forms
accessing the same limited soil moisture, trees would be at a significant
disadvantage to shrubs (an overall shorter conducting system). On this
point Balding and Cunningham (1974) showed that plant height became
progressively shorter in Acacia constricta (tree) and Larrea tridentata (shrub)
as drought severity increased along an aridity gradient. Stem xylem cavitation
events would be expected to proceed from the stem tip where transpiration
is occurring toward basal regions, creating progressive stem dieback under
conditions of increasing drought duration. This may in part be the explanation for why suffrutescent growth (many stems emerging from a common
root base) is so common among aridland shrubs.
Life-form diversity of perennial species is positively related to climatic
diversity. Cody (1989) showed that life-form diversity among Sonoran Desert
sites increased under extreme conditions - conditions likely to reflect both
extremes in drought and a high year-to-year variability in precipitation.
Shmida and Burgess (1988) found that biseasonality of precipitation was the
major driving factor accounting for the increased growth-form diversity
in the Sonoran Desert. Biseasonal precipitation was also correlated with
these same life-form patterns within the Great Basin and Colorado Plateau
(Comstock and Ehleringer 1992). Together, these data suggest that in North
American deserts it is the diversity of incoming moisture sources that drives
the evolution of differences in life-form.
The coefficient of variation is a measure of unpredictability (Sect. 18.2.1)
and an unpredictable environment will preclude the evolution of perennial
plants dependent on frequent rainfall episodes or not capable of persisting
through extended drought periods. The loss of constancy (the degree to
which conditions are similar from year to year) in drier sites should have a
direct impact on the distribution of life-forms along an aridity gradient.
Hastings and Turner (1965a) examined the amounts and variability of winter
and summer precipitation in Baja California. They noted that the variability
in precipitation corresponded to the distribution of vegetation types described
by Shreve (1934) and Shreve and Wiggins (1964). After Colwell (1974)
showed that predictability and constancy were mathematically related,
Schaffer and Gadgil (1975) then followed with a prediction of the distribution
abundances of annual versus perennial life-forms as predictability of rainfall
decreased. The data of Harper (unpub!' data cited in Schaffer and Gadgil
1975) supported the prediction that annuals become an increasing fraction of
the total flora as predictability of that precipitation decreases. What these
data sets did not reveal was any information on the distribution of photosynthetic pathways among these annuals. Such information can be extracted
from Shreve and Wiggins (1964), who noted that there were distinct winterand summer-annual floras, even within a particular region. It turns out that
winter annuals have C 3 photosynthesis exclusively, while summer annuals