Agriculture Reference
In-Depth Information
infection, determinate nodules) suggest that burkholderias can infect legumes and in-
duce nodulation in all the ways shown for classical rhizobia. Further, plants nodulated
by
-rhizobia may be herbaceous, shrubby, trees, annual or perennial. At the moment
all reports are from tropical or warm temperate areas and information to date indicates
that
-rhizobia (e.g. Elliott et al., 2007a).
There may be considerable geographical variation in the occurrence of
Burkholderia
species in nodules. For example, Parker (2008) looked at bacteria in nodules of 24
species from 18 legume genera, including all sub-families, in Barro Colorado Island,
Panama and found the majority to be bradyrhizobia, with the next most abundant
being
Burkholderia
, but the latter were only found in plants from Mimosoideae. There
may also be within genus variation, for example in the papilionoid
Lebeckia
, where
most species nodulate with a variety of
-rhizobia have acquired
nod
genes from local
-rhizobia, but those from section Lebeckia
may be nodulated by
Burkholderia
(Phallane et al., 2008)
4.3 Other bacterial nodule occupants
Nodules may house various bacteria, not all of which are able either to fix nitrogen
or to induce nodules. It is not sufficient to isolate a bacterium from a nodule and
assume that it can nodulate the host of isolation, a fact that has led to a number of
unfortunate errors in recent publications, even in prestigious journals such as
Nature
(e.g. Moulin et al., 2001). In one case the bacteria were from a nodulating strain,
but not for the plant from which it was allegedly isolated. It is essential that Koch's
postulates be established before claims are made, particularly if the bacteria concerned
are not known to be nodulating forms. Benhizia et al. (2004) isolated several genera
of
-Proteobacteria from nodules of
Hedysarum
species and implied that these could
nodulate their hosts of isolation (indeed the title of the paper said so), but no evidence of
this was given, and in a later paper they found that these
-Proteobacteria were unable
to nodulate their hosts (Muresu et al., 2008). These workers had a particularly difficult
situation in that the rhizobia in their active nodules were very difficult to culture. Chou
et al. (2007) isolated a new species of the
-Proteobacterial genus
Labrys
,
L. neptuniae
,
from nodules of
Neptunia oleracea
, but showed that this bacterium could not nodulate
its host of isolation. To date, there is no good evidence that
-Proteobacteria can
nodulate legumes. On the other hand, non-nodulating bacteria from various genera
are frequently isolated from nodules. In a detailed study of bacteria isolated from
soybean nodules in China, 98 strains of non-nodulating forms from the
,
,and
-Proteobacteria were found (Li et al., 2008). Some of these could produce indole acetic
acid, could solubilise mineral phosphate and fix nitrogen, with their
nifH
genes being
very similar to those of the nodulating bacterium
Bradyrhizobium japonicum
. Although
these non-nodulating bacteria had the potential to aid plant growth, Li et al. (2008)
were unable to obtain evidence that they actually did so. These and other studies,
not cited, give very strong evidence that legume root nodules may contain a range of
bacteria that are unable to nodulate their hosts. How they get into the nodules, and
what, if anything, they do there (apart from having a nice culture medium to live in),
is not clear. Whether they are more common in wild legumes than cultivated ones, as
