164
M.T.K. Arroyo et al.
be on Australia, where well-developed altitudinal temperature corridors
are less conspicuous. Of course, verification of the existence of Rapoport's
rule under these conditions becomes a necessary first step.
The tendency for few species per genus seen in southern rainforest
also seems to characterize North American and Tasmanian rainforest
(Tasmania, Jarman et al., 1991; Alaback, this volume) . These are interesting findings . Documented cases of hybridization are rare in southern
rainforest. For the latter, this is counterintuitive, considering the importance of large-scale disturbance due to landslides and volcanic activity
(Veblen & Ashton, 1978; Veblen, Donoso, Schlegel, & Escobar, 1981).
If anything, the South American environment should provide an ideal
setting for hybridization and reticulate evolution. We hypothesize that
lack of hybridization is a direct consequence of a generically rich flora,
with low number of species per genus. Together, these ensure little
opportunities ·for cogeneric encounters, thus minimising hybridization.
Interestingly, the best-documented cases of hybridization in forest species
in southern South America are in the genus Nothofagus (Donoso, 1984b,
1987; Donoso & Atienza, 1983), which is also thelargest tree genus in the
southern South American forest flora.
As a general theme in rainforest research, it would be useful to develop
testable hypotheses for the causes of low numbers of species per genus.
Steven's (1989) tries to explain low species.numbers in general at high
latitudes in a biogeographic framework. It seems particularly relevant in
this context that few of the "taxonomic relics" in rainforest are relic in
the ecological sense. Some in fact are dominant (e.g., Aextoxicon, Fitzroya,
Pilgerodendron, Laureliopsis) and many are fairly common (e.g.,
Embothrium, Saxegothaea) . Such wide geographic distributions in
monotypic genera would seem to indicate that the forcing factor for low
number of species per genus is not that such genera are out of phase with
their present environment, but that they are prevented in some way from
speciating.
Genetic constraints could have developed in the Pleistocene in some
long-lived species like Fitzroya cupressoides. Cool, Power, and Zavarin
(1991) recently showed that this species is invariant for monoterpenes and
appears to have been bottlenecked. A low level of genetic variation has
also been demonstrated in Lagarostrobos [ranklinii (huon pine), an
endemic Tasmanian rainforest conifer (Shapcott, 1991). Alternatively, in
some taxa, a low level of speciation might reflect interplay between
population density and plant breeding system (Arroyo et al., 1993) and
some of the more general causes of species numbers at different latitudes
offered by Stevens (1989). Riveros (1991) has recently shown that many
rainforest trees are dioecious or genetically self-incompatible (M. Riveros,
unpublished data). Such breeding systems can slow the rate of population
differentiation, although populations may have high levels of hetero-
Précédent

- 185/430

Suivant