8. Phytogeographic Relationships and Regional Richness Patterns
161
trend occurs along some elevational gradients (Stevens, 1992). The same
author illustrated that the latitudinal trend in the Northern Hemisphere
is paralleled by an increase in the temperature extremes an organism
must experience throughout its lifetime. Stevens (1989) sought a causal
explanation of Rapoport's rule from the latter correlation, whereby highlatitude organisms, in being adapted to a wider range of temperature
conditions, should be able to occupy larger latitudinal ranges and will be
less likely to be habitat specialists. The reduced range of temperature
conditions experienced by organisms in low-latitude environments promotes the evolution of habitat specialization. Stevens (1989)further argued
that Rapoport's rule could be a causal factor of the typical latitudinal
gradient in species richness. As a particular species range increases, space
for evolution of other species is preempted. We have looked for these
correlations in the trees species of the cool temperate rainforest of southern
South America, in view of the absence of an increase in annual temperate
range along the entire latitudinal extent of rainforest, discussed earlier.
Stevens (1989) demonstrated latitudinal increase in temperature
extremes by comparing the absolute maximum and minimum temperatures
recorded. This kind of data was unavailable to us for South America.
Moreover, use of these values seems dubious, since a brief temperature
pulse is unlikely to have any long-standing biological impact. In any case,
to satisfy ourselves that the annual temperature ranges as used here show
the same tendency, we compared differences in .absolute extremes with
the two measures used here for the Northern Hemisphere stations shown
in Figure 8.3, obtaining highly positively correlations (r = 0.811; p <
0.001 for the range based on means; r = 0.829; p < 0.001 for range based
on highest mean maximum and lowest mean minimum). The slope of the
curve was 1.09 in the second case and 1.3 in the first, indicating very
faithful representation in the second case.
Figure 8.7 shows the mean latitudinal range of tree species occurring
at two degree latitudinal intervals from 38° to 55°S in southern South
America . At 38°S,the mean latitudinal range was the equivalent to
around 14°, whereas by 55°S, it has increased to around 21°, indicating a
substantial increase in range size for trees over the latitudinal gradient
studied. It can be seen that this increase persists, in spite of no evident
increase in annual temperature range as we have measured it.
Discussion and Conclusions
This paper has shown that in spite of the evident isolation of southern
South American rainforest from other wet forest habitats. in South
America today, there seem to be considerably more genera and species,
161
trend occurs along some elevational gradients (Stevens, 1992). The same
author illustrated that the latitudinal trend in the Northern Hemisphere
is paralleled by an increase in the temperature extremes an organism
must experience throughout its lifetime. Stevens (1989) sought a causal
explanation of Rapoport's rule from the latter correlation, whereby highlatitude organisms, in being adapted to a wider range of temperature
conditions, should be able to occupy larger latitudinal ranges and will be
less likely to be habitat specialists. The reduced range of temperature
conditions experienced by organisms in low-latitude environments promotes the evolution of habitat specialization. Stevens (1989)further argued
that Rapoport's rule could be a causal factor of the typical latitudinal
gradient in species richness. As a particular species range increases, space
for evolution of other species is preempted. We have looked for these
correlations in the trees species of the cool temperate rainforest of southern
South America, in view of the absence of an increase in annual temperate
range along the entire latitudinal extent of rainforest, discussed earlier.
Stevens (1989) demonstrated latitudinal increase in temperature
extremes by comparing the absolute maximum and minimum temperatures
recorded. This kind of data was unavailable to us for South America.
Moreover, use of these values seems dubious, since a brief temperature
pulse is unlikely to have any long-standing biological impact. In any case,
to satisfy ourselves that the annual temperature ranges as used here show
the same tendency, we compared differences in .absolute extremes with
the two measures used here for the Northern Hemisphere stations shown
in Figure 8.3, obtaining highly positively correlations (r = 0.811; p <
0.001 for the range based on means; r = 0.829; p < 0.001 for range based
on highest mean maximum and lowest mean minimum). The slope of the
curve was 1.09 in the second case and 1.3 in the first, indicating very
faithful representation in the second case.
Figure 8.7 shows the mean latitudinal range of tree species occurring
at two degree latitudinal intervals from 38° to 55°S in southern South
America . At 38°S,the mean latitudinal range was the equivalent to
around 14°, whereas by 55°S, it has increased to around 21°, indicating a
substantial increase in range size for trees over the latitudinal gradient
studied. It can be seen that this increase persists, in spite of no evident
increase in annual temperature range as we have measured it.
Discussion and Conclusions
This paper has shown that in spite of the evident isolation of southern
South American rainforest from other wet forest habitats. in South
America today, there seem to be considerably more genera and species,