Agriculture Reference
In-Depth Information
three different areas where the information is sufficiently robust. There is no complete
recent flora of China, although there is great international effort underway to produce
one. There is an excellent checklist of legumes of South Asia (Kumar & Sane, 2003)
and of West Asia (Lock & Simpson 1991), but it has not been possible to sort these into
vegetation types. The easternmost of the three examples in Table 2.3 is Borneo followed,
in a westerly direction, by the Guineo Congolean region and then a composite South
American example. Table 2.3A is derived from the database at the University of Leiden
after deleting all those species that appear to have been introduced. Table 2.3B and C
are for woody species, including lianes. Data for the Guineo-Congolean forests, which
are found throughout equatorial Central and West Africa, are taken from Lock (1989),
incorporating taxonomic changes given in Lewis et al. (2005). Major South American
tropical rain forests of the Guyanas (including Iwokrama) and the Reserva Ducke in
the Terra Firme region of Amazon Brazil are summarised in Table 2.3C. These data
are largely taken from Clarke et al. (2001), but the flora of the Reserve Ducke is also
comprehensively covered in the monograph edited by Ribeiro et al. (1999).
It must be emphasised that there are major differences within the African and South
America forest areas. On the whole, there is less diversity in African than Amazonian
rain forests, and this has been linked to climate variables (Parmentier et al., 2007).
A separate, but notable difference is that whereas there have been many years of
taxonomic study in both regions, in South America, there has also been an emphasis
on looking for nodulation - see for example Faria et al. (1989), Moreira et al. (1992),
Roggy & Prevost (1999) and Perreijn (2002). In Africa, Corby's work (e.g. 1988) alone
is pre-eminent. Thus in Africa and in Borneo there are many genera whose nodulation
status is unknown, whereas in South America there are few such genera and none in
the regions covered in Table 2.3C. However, even in Africa, many of the genera for
which there are no reports, particularly in Caesalpinioideae (Sprent, 2007) are, from
their taxonomic affinities, unlikely to be able to nodulate. More than half of the legume
genera in the African forest are caesalpinioid, whereas less than half are in the South
America forest. Also, fewer of the caesalpinioid legumes in the African forests are able
to nodulate. Is this reflected in the contribution of nitrogen fixation to these ecosystems?
A very surprising finding of a trip to Cameroon in 2000 was the extent of nodulation on
the roots of mature trees and lianas, such as species of
Pterocarpus
and
Mucuna
(Sprent,
2001), often tens of metres from the parent plant. This observation was counter to the
usually perceived wisdom that mature trees have a very efficient system for recycling
nitrogen and only a few nodules are needed to top up any losses. For the same reason,
it is usually recommended that to look for nodulation ability in woody plants, young
seedlings should be examined. Sprent (2001) suggested that the profusely nodulated
plants may fix enough nitrogen for the ecosystem and that this is transported between
plants by mycorrhizas. This suggestion has not been tested in a rain forest, but there
is now increasing evidence for nitrogen transfer in both directions between fixing and
non-fixing plants (He et al., 2003). This has management implications, if the nodulated
trees are selectively logged. It is fair to say that we have no idea how much nitrogen
is fixed by legumes in African rain forests. On the other hand, there is evidence from
South American forests of all the regions covered in Table 2.3C, as well as in the wider
Amazonian region, that significant nitrogen fixation can occur (Martinelli et al., 1992;
Moreira et al., 1992; Roggy & Prevost, 1999), although there are many local variations.
