• Leguminosae;
  • nitrogen fixation;
  • nodulation identification;
  • nodulation status;
  • nodule-like structures;
  • tropical rainforest legumes

When is a nodule not a nodule?

  1. Top of page
  2. When is a nodule not a nodule?
  3. Can legume taxonomy help?
  4. The significance of nodulation in African rainforest legumes: a longitudinal perspective
  5. Acknowledgements
  6. References

In a recent paper, Diabate et al. (2005) rightly pointed out that, compared with South America, the tropical rainforest legumes of Africa have been little studied for their ability to nodulate and fix nitrogen. These authors examined 156 species in six different forests and reported on their nodulation. Overall their data are consistent with the known distribution of nodulating behaviour in the Leguminosae. Some of their new reports are on species of genera widely reported as nodulating. Others are on new genera, from sections of the Leguminosae that include nonnodulating genera. Because the criteria on which their nodule identification was based were not given, it is difficult to evaluate these data.

The term ‘nodule’ is widely used in geology, medicine, botany and other disciplines. With respect to legumes, its use is usually confined to swellings on roots (occasionally stems) that house unicellular bacteria with the ability to fix atmospheric nitrogen. Many woody plants, including legumes, may have swellings on roots that superficially appear to be nodules, but are not. This has led to false reports of nodulation in the literature. In a detailed survey of leguminous trees in a tropical rain forest in Guyana, Perreijn (2002) divided species into three categories: those with nitrogen fixing nodules, those with nodule-like structures on roots and those without either. In a careful examination of the nodule-like swellings on the roots of the caesalpinioid species Eperua falcata Aubl. and Mora excelsa Benth., no nitrogenase activity was found, nor could nitrogen-fixing bacteria be isolated from them. Photographs of these structures are shown in Sprent (2001). Small, nodule-like structures were also seen on Chamecrista adiantifolia (Benth.) Irwin & Barneby, by Perreijn (2002), a species that is known to nodulate (Corby, 1988). The caesalpinioid genus Chamaescrista has at least 50 known nodulating species (Sprent, 2001), and for a number of these there is both structural (Naisbitt et al., 1992) and functional (Sprent et al., 1996) evidence that they fix nitrogen. It is thus possible that the structures recorded by Perreijn (2002) on C. adiantifolia were young or ineffective nodules, as she suggested. No structural studies were performed on them. As has been pointed out on a number of occasions (e.g. de Faria et al., 1989; Sprent, 2001), it is very difficult to be certain from field specimens that a swelling on a root is a nodule. It can be equally easy to decide on external appearance that some structures that are nodules are not, because their morphology and, in some cases, their internal structure are unlike those of more familiar agricultural species (Sprent, 2001).

In order to confirm that a species nodulates, one or more of a number of lines of additional evidence are needed. Field observations should ensure that the nodulated roots were attached to the putative host plant; position of nodules on the root systems and their external shape should be recorded (e.g. de Faria et al., 1984; Corby, 1988). In many cases, nodule morphology, position on roots (whether or not associated with lateral roots) and nodule structure are taxonomic characters (Corby, 1988; Sprent, 2000). In addition, seedlings should be grown in soil from their native habitat and observation of development of swellings on roots should be made, and this is planned by Diabate et al. (2005) for their material (A. Galiana, INRA, Montpellier, pers. comm.). Following this, or on field specimens, evidence of nitrogenase activity (usually by the acetylene reduction assay) and/or simple sectioning to show internal colouration of nodules (usually pink when active, due to haemoglobin, but not always, as some nematode galls are also pink) and location of infected tissue (central in legumes) are simple checks that can be followed up with more detailed microscopy, including immunolocalization of nitrogenase and other key enzymes (Chen et al., 2003).

Not all laboratories in countries with the most interesting legume trees have the facilities to carry out a full suite of tests for true nodulation and nitrogen fixation, but a subset of morphology and simple light microscopy, coupled with a knowledge of host plant taxonomy, can help to support visual observations until more detailed studies on fixed material and isolation of bacteria can be carried out. Fortunately, isolation can be achieved quite simply from nodules dried over a suitable desiccant and light microscopy on material immersed in old-fashioned fixatives such as 95% ethanol.

In the same way that observation of swellings on roots is not sufficient to establish nodulation, isolation of particular bacteria from nodules is not sufficient to prove that these bacteria can induce nodules. Isolated bacteria should be re-inoculated on to their putative hosts and shown to produce nodules, as Diabate et al. (2005) have done for species of Albizia, Erythrophleum and Millettia. This basic bacteriological technique, which has been taught at undergraduate level for nearly a century, has been used by some (e.g. Sy et al., 2001) but not all (e.g. Moulin et al., 2001) recent publications reporting new genera of bacteria that may nodulate legumes. Modern methods, such as tagging bacteria with green fluorescent protein, make observations of nodulation processes much easier (Chen et al., 2003).

Can legume taxonomy help?

  1. Top of page
  2. When is a nodule not a nodule?
  3. Can legume taxonomy help?
  4. The significance of nodulation in African rainforest legumes: a longitudinal perspective
  5. Acknowledgements
  6. References

The family Leguminosae (Fabaceae) is generally subdivided into three subfamilies, Caesalpinioideae, Mimosoideae and Papilionoideae, and the extent of nodulation varies widely within these, from uncommon, through common to very common, within subfamilies, considerable correspondance with tribal affinities. Thus it is particularly important to be careful when reporting nodulation in groups where it is uncommon. Diabate et al. (2005) took this into consideration when discussing some of their observations, pointing out that the monotypic caesalpinioid genus Chidlowia (reported to nodulate) is in the tribe Caesalpinieae, where seven of the eight confirmed nodulating caesalpinioid legumes are placed. However, tribe Caesalpinieae has about 50 genera, for some of which there are no reports on nodulation status, others for which there are unconfirmed reports and yet others with several reports of nonnodulation. The tribe also includes Eperua and Mora, discussed above, as well as the only confirmed African caesalpinioid nodulating genus, Erythrophleum. It will be particularly interesting to see whether the field observations on Chidlowia are confirmed and whether any nodules produced have the characteristic primitive structure associated with nodules from all tree species in the Caesalpinioideae that have been studied (Sprent, 2001).

Within the subfamily Mimosoideae, nodulation is much more common, but not universal. Genera considered to be basal are trees and are usually included in tribes Parkieae and Mimoseae. However, Parkieae, which contains only two genera (Parkia and Pentaclethra) is now considered to be polyphyletic (Luckow et al., 2003), with Parkia (nonnodulating) more closely related to members of tribes Ingeae and Mimosae than to Pentaclethra (nodulating in Africa, but not South America) (Sprent, 2001). Relations within tribe Mimoseae have also been revised by Luckow et al. (2003), and this is of interest with respect to nodulation generally and to the reports of Diabate et al. (2005), who give new positive reports of nodules on species of Aubrevillea, Calpocalyx, Tetrapleura and Xylia. There are no published molecular data on Aubrevillea, which was tentatively put in a group of its own within the Mimoseae by Luckow et al. (2000). This genus of two species is endemic to the Guineo–Congolean forest, so it will be interesting to see further information on its nodulation status. Genera Calpocalyx, Tetrapleura and Xylia all have negative reports on nodulation in the literature. On molecular and other criteria, they appear to be fairly closely related to each other and also to the genus Adenanthera, which is generally considered not to nodulate (Sprent, 2001; Luckow et al., 2003). Again, confirmatory data are needed to verify nodulation in these genera.

Within the subfamily Papilionoideae, nodulation is recorded by Diabate et al. (2005) for Amphimas, a genus found not to be nodulated in Cameroon (D. Nwaga, University of Yaounde 1, pers. comm.; see Sprent, 2001). This genus is currently placed in the polyphyletic tribe Sophoreae, which contains both nodulated and nonnodulated genera (Sprent, 2001). However, Amphimas is considered to be basal within this tribe and the subfamily (Pennington et al., 2001). Recent molecular studies on base papilionoid genera are grouping together several genera from various tribes, including Sophoreae, which appear unable to nodulate (Pennington et al., 2001; Sprent, 2001). Molecular data are not yet available for Amphimas and there are no reports on nodulation for a number of other genera in this basal group. More information on these is needed to assess any relationship between nodulation criteria and molecular and other taxonomic characters. The utility of nodulation characters in legume taxonomy has been discussed by Sprent (2000), and in a review of the dalbergioid group of papilionoid legumes, nodulation characters were among those most strongly related to molecular evidence (Lavin et al., 2001).

The significance of nodulation in African rainforest legumes: a longitudinal perspective

  1. Top of page
  2. When is a nodule not a nodule?
  3. Can legume taxonomy help?
  4. The significance of nodulation in African rainforest legumes: a longitudinal perspective
  5. Acknowledgements
  6. References

As Diabate et al. (2005) point out, a large proportion of trees in humid forests of West Africa are legumes and of these many are important timber species. Sprent (2001) grouped genera listed from Guineo–Congolean forests by Lock (1989) with some updates on taxonomy from later work and from the International Legume Database & Information Service (ILDIS) database ( These data are summarized and extended in Table 1. Clearly there is a lot of work to be done to check the nodulation status of genera in these forests. Even then, as Diabate et al. (2005) and others have reported, not all species known to nodulate will do so in a particular area, owing to factors such as soil type and presence of suitable endophytic bacteria.

Table 1.  Woody legumes (trees and lianas) of the Guineo–Congolean forests and their putative nodulation status
Subfamily, tribenodulatednonnodulatedunknown
  1. These data are summarized and modified from Sprent (2001). The numbers of genera and their tribal affinities are currently under review.

Caesalpinieae 1 1 4
Cassieae 0 2 2
Cercideae 0 1 1
Detarieae 02319
Acacieae 1 0 0
Ingeae 2 0 0
Mimoseae 2 5 4
Parkieae 0 1 0
Aeschynomeneae 1 0 0
Dalbergieae 3 0 0
Millettieae10 0 3
Phaseoleae 6 0 1
Sophoreae 1 3 5
Swartzieae 1 2 0

Taking a longitudinal perspective of tropical rain forests, those in Asia appear to have a relatively low proportion of legumes, but both nodulated and nonnodulated species are found there (Sprent, 2001). African forests have a high proportion of legumes, but many of them are caesalpinioid, base mimosoid and papilionoid and lack the ability to nodulate (Table 1). In South America, legumes are also very common, but there appears to be a higher proportion of nodulating species from all subfamilies (Sprent, 1999, 2001). The occurrence and significance of mycorrhizas in African rainforests has been widely studied, especially with respect to phosphorus dynamics (e.g. Newbery et al., 1997). However, their nitrogen dynamics has been largely ignored. Does the small proportion of nodulated trees play an important role? Recent work from the Cameroon, as reported briefly by Sprent (2001, 2002), suggests that it may. In the Bayang-Mbo forest, large numbers of nodules were found in the litter layer and nodulated roots were often seen growing up trunks of trees of nonnodulating legumes and other species (Fig. 1a,b). These were traced for many metres (up to at least 30 m in some cases) until they could be connected to their parent trees or lianas (Fig. 1c). Surpisingly, mature trees were found to have copiously nodulated roots (Fig. 1c). It is generally considered that large trees recycle much of their nitrogen and therefore have little need for nodulation. This is one reason why it is recommended (e.g. de Faria et al., 1989) that it is best to look for nodulation in seedlings rather than in mature trees. Nodules found in this Cameroon forest were of a variety of types, but fortunately these included classical desmodioid and aeschynomenoid forms (Corby, 1988; Sprent, 2001), known to be associated with particular genera. Desmodioid nodules are more or less spherical, with prominent lenticels and pronounced internal pink colouration. Nodules of this type were traced to a liana (probably a species of Mucuna), with a stem that went vertically up to the canopy (Fig. 1d). Aeschynomenoid nodules are oblate and formed in the axils of lateral roots; these were traced back to a large tree of Pterocarpus, a genus known to have this type of nodule (Corby, 1988; Sprent, 2001). In the same area of forest, most of the legume trees were nonnodulated caesalpinioid species, often from tribe Detarieae (Table 1). In other parts of the forest, nodulated species of papilionoid genera Baphia, Millettia and Erythrina were found, the former two also reported from Guinea by Diabate et al. (2005).


Figure 1. Nodulated legumes in the Bayang-Mbo Forest, Cameroon. (a) A nodulated root climbing up the trunk of a nonnodulating caesalpinioid legume – the spherical nodule has a lenticel (seen as a stripe) typical of the desmodioid type of nodule found in papilionoid tribes Phaseoleae and Desmodieae (now usually grouped as one tribe). (b) A continuous root system, with desmodioid nodules attached, within the litter layer was traced approximately 25 m back to the base of a liana where many more nodules were attached (c). The stem of this liana went vertically up to the forest canopy (d).

Download figure to PowerPoint

Apart from the few reports of nodulation that need to be confirmed, Diabate et al. (2005) generally found lack of nodulation in large numbers of woody members of caesalpinioid tribes Amherstieae, Cassieae, Cercideae and Detarieae. These observations, together with those from Cameroon, cited above, raise the question as to whether a few copiously nodulated species can fix enough nitrogen for the needs of the whole forest, supplementing that which is recycled. If this is the case, it raises another possibility, that mycorrhizas may have a function in transferring nitrogen between species. This possibility is supported by recent work of He et al. (2003) showing that there can be two-way transfer of nitrogen between nodulated and nonnodulated woody species. It has important practical implications. If nodulated trees are selectively logged, which may be the case (for example, for Pterocarpus, which is prized for its reddish-coloured wood), then the nutrient dynamics of the whole forest could be at risk.

Clearly, as Diabate et al. (2005) stress, we need to know much more about the nodulation status of West African legumes if forests are to be properly understood and managed. This will require much careful work on the part of many people, particularly those who live in the countries concerned.


  1. Top of page
  2. When is a nodule not a nodule?
  3. Can legume taxonomy help?
  4. The significance of nodulation in African rainforest legumes: a longitudinal perspective
  5. Acknowledgements
  6. References

I would like to thank Barbara Mackinder (RBG Kew, UK), Nwaga Dieudonné and Kiam Angele (University of Yaounde 1, Cameroon) for permission to quote our joint work, initially reported in Sprent (2001), and the Royal Society of Edinburgh for a travel grant enabling me to visit Cameroon. I am also grateful to Antoine Galiana, the corresponding author of Diabate et al. (2005), for email correspondence on their paper, indicating how this work will be taken forward.


  1. Top of page
  2. When is a nodule not a nodule?
  3. Can legume taxonomy help?
  4. The significance of nodulation in African rainforest legumes: a longitudinal perspective
  5. Acknowledgements
  6. References
  • Chen WM, James EK, Prescott AR, Kierans M, Sprent JI. 2003. Nodulation of Mimosa spp. by the β-proteobacterium Ralstonia taiwanensis. Molecular Plant–Microbe Interactions 16: 10511061.
  • Corby HDL. 1988. Types of rhizobial nodule and their distribution among the Leguminosae. Kirkia 13: 53123.
  • Diabate M, Munive A, De Faria SM, Ba A, Dreyfus B, Galiana A. 2005. Occurrence of nodulation in unexplored leguminous trees native to the West African tropical rainforest and inoculation response of native species useful in reforestation. New Phytologist 166: 231239.
  • De Faria SM, Franco AA, De Jesus RM, De Menandro MS, Baitello JB, Mucci ESF, Dobereiner J, Sprent JI. 1984. New nodulating legume trees from South-East Brazil. New Phytologist 98: 317327.
  • De Faria SM, Lewis GP, Sprent JI, Sutherland JM. 1989. Occurrence of nodulation in the Leguminosae. New Phytologist 111: 607619.
  • He X-H, Critchley C, Bledsoe C. 2003. Nitrogen transfer within and between plants through common mycorrhizal networks (CMNs). Critical Reviews in Plant Sciences 22: 531567.
  • Lavin M, Pennington RT, Klitgaard BB, Sprent JI, De Lima HC, Gasson PE. 2001. The Dalbergioid legume (Fabaceae): delimitation of a pantropical monophyletic clade. American Journal of Botany 88: 503533.
  • Lock JM. 1989. Legumes of Africa: A Check-List. Kew, UK: Royal Botanic Gardens.
  • Luckow M, Miller JT, Murphy DJ, Livshultz T. 2003. A Phylogenetic analysis of the Mimosoideae (Leguminosae) based on chloroplast DNA sequence data. In: KlitgaardBB, BruneauA, eds. Advances in Legume Systematics, Part 10, Higher Level Systematics. Kew, UK: Royal Botanic Gardens, 197220.
  • Luckow M, White PJ, Bruneau A. 2000. Relationships Among the Basal Genera of Mimosoid Legumes. In: HerendeenPS, BruneauA, eds. Advances in Legume Systematics , Part 9. Kew, UK: Royal Botanic Gardens, 165180.
  • Moulin L, Munive A, Dreyfus B, Boivin-Masson C. 2001. Nodulation of legumes by members of the β-subclass of Proteobacteria. Nature 411: 948950.
  • Naisbitt T, James EK, Sprent JI. 1992. The evolutionary significance of the genus Chamaecrista as determined by nodule structure. New Phytologist 122: 487492.
  • Newbery DMcC, Alexander IJ, Rother JA. 1997. Phosphorus dynamics in a lowland African rain forest, the influence of ectomycorrhizal trees. Ecological Monographs 67: 367409.
  • Pennington RT, Lavin M, Ireland H, Klitgaard BB, Preston J, Hu J-M. 2001. Phylogenetic relationships of basal papilionoid legumes based upon sequences of the chloroplast trnL intron. Systematic Botany 26: 537556.
  • Perreijn K. 2002. Symbiotic Nitrogen Fixation by Leguminous Trees in Tropical Rain Forest in Guyana. In: Tropenbos-Guyana Series 11. Guyana, Georgetown: Tropenbos.
  • Sprent JI. 1999. Nitrogen fixation and growth of non-crop species in diverse environments. Perspectives in Plant Ecology, Evolution and Systematics 2: 149162.
  • Sprent JI. 2000. Nodulation as a Taxonomic Tool. In: HerendeenPS, BruneauA, eds. Advances in Legume Systematics., part 9. Kew, UK: Royal Botanic Gardens, 2144.
  • Sprent JI. 2001. Nodulation in Legumes. Kew, UK: Royal Botanic Gardens.
  • Sprent JI. 2002. Knobs, knots and nodules – the renaissance in legume symbiosis research. New Phytologist 153: 26.
  • Sprent JI, Geoghegan IE, Whitty PW, James EK. 1996. Natural abundance of 15N and 13C in nodulated legumes and other plants in the Cerrado and neighbouring regions of Brazil. Oecologia 105: 440446.
  • Sy A, Giraud E, Jourand P, Garcia N, Willems A, De Lajudie P, Prin Y, Neyra M, Gillis M, Boivin-Masson C, Dreyfus B. 2001. Methylotrophic Methylobacterium nodulate and fix nitrogen in symbiosis with legumes. Journal of Bacteriology 183: 214220.