Agriculture Reference
In-Depth Information
described by Doyle and Luckow (2003). More recently Schrire et al. (2005) pointed out
some anomalies in legume distribution that did not support the boreal hypothesis. In-
stead they suggested that legumes could hop across islands or otherwise be carried by
wind or sea water (see also Renner, 2004a) or by extreme weather events (Sprent, 2007).
Schrire et al. (2005) further carried out a very comprehensive analysis of where extant
legume genera are found. As a result they suggested that the current distribution may
be more related to particular biomes (see Chapter 2), where genera and species have ar-
rived by dispersal, than to vicariance, where species evolve following geographic isola-
tion. In addition they suggested that rather than evolving in the humid tropics, legumes
arose in a semi-arid area just north of the Tethys Sea, which separated the two major
land masses existing at that time. The earliest legumes were adapted to semi-arid con-
ditions (the succulent biome) and subsequently spread into arid, fire-tolerant areas (the
grass biome), into humid areas (rain forest biomes) and into north and south temper-
ate areas (Chapter 2). These movements occurred while land masses were also moving
and were related to major climatic shifts. Schrire et al. (2005) list a set of characteristics
found in fossil legumes that are consistent with a semi-arid environment and that are
found in extant legumes in such areas. These include compound leaves, which can be
shed in part or completely during periods of drought. In extant SDTF there is a mixture
of nodulated and non-nodulated legumes, as there appears to have been in the original
biome north of the Tethys Sea. An interesting question is when legumes diversified
into the grass biome, which could have been comparatively recently, since C
4
grasses
evolved less than 10 Ma (Sprent, 2007). Whenever it was, few non-nodulating genera
appear to have made this move (Chapter 2). This contrasts with rain forests, where
there are numerous non-nodulating genera, with major groups still actively evolving.
As mentioned in Chapter 2, virtually all legumes in temperate biomes are nodulated.
3.2.1 Madagascar as a special case
In a comprehensive review of vicariance versus dispersal as ways of explaining the
present diversity of animals and plants in Madagascar, Yoder and Nowak (2006) point
out the many sister groups between Africa and Madagascar. As far as higher plants
are concerned the picture is complex. It is generally agreed that Madagascar separated
from the main Indian land mass about 88 Ma. Molecular analysis of the widespread
Asian ectomycorrhizal family Dipterocarpaceae suggests that extant species of this
and the derived family Sarcolaenaceae (endemic to Madagascar) arose from ancestors
present before the separation (Ducousso et al., 2004). However, as discussed above,
legumes evolved after the separation of these two land masses, making the situa-
tion more complex. Fortunately there is a comprehensive work on the legumes of
Madagascar (Du Puy et al., 2002), and this will be used here to highlight some of the
problems. Table 3.2 summarises the native legume species found there, most of which
are endemic. All the genera specified in the comments column apart from
Viguieranthus
(which is endemic to Madagascar) and
Mimosa
are well represented in Africa and have
probably arrived from there. Indeed, as has been suggested for members of family
Melastomataceae (Renner, 2004b), each genus may have made several separate trips.
For example, four such transfers have been suggested for
Indigofera
, possibly in the
