Protoplasma

DOI 10.1007/s00709-015-0922-2

ORIGINAL ARTICLE

Symbiotic functioning, structural adaptation, and subcellular

organization of root nodules from Psoralea pinnata (L.) plants
grown naturally under wetland and upland conditions in the Cape
Fynbos of South Africa
Sheku A. Kanu 1 & Felix D. Dakora 2

Received: 11 June 2015 / Accepted: 1 December 2015

# Springer-Verlag Wien 2015

Abstract In the Cape Fynbos of South Africa, Psoralea

pinnata (L.) plants occur naturally in both wetland and welldrained soils and yet effectively fix N2 under the two contrasting conditions. In this study, nodule structure and functioning
in P. pinnata plants from the two habitats were evaluated using
light and transmission electron microscopy (TEM), as well as
the 15N natural abundance technique. The results showed that,
structurally, fully developed P. pinnata nodules were spherical
in shape with six components (namely, lenticels, periderm,
outer cortex, middle cortex, inner cortex, and a central
bacteria-infected medulla region). Morphometric analysis revealed 44 and 84 % increase in cell area and volume of wetland nodules compared to those from upland. The percentage
area of nodules occupied by the middle cortex in wetland
nodules was twice that of upland nodules. As a result, the size
of the medulla region in wetland nodules was significantly
reduced compared to upland nodules. Additionally, the average area of medulla occupied by intercellular air spaces in
wetland nodules was about five times that of upland nodules
(about 431 % increase in wetland over upland nodules). TEM
data also showed more bacteroids in symbiosomes of upland
nodules when compared to wetland nodules. However, isotopic analysis of above-ground plant parts revealed no differences in symbiotic parameters such as N concentration, 15N

Handling Editor: Adrienne R. Hardham

* Felix D. Dakora
dakorafd@tut.ac.za
1

Crop Sciences, Tshwane University of Technology, Private Bag

X680, Pretoria 0001, South Africa

Chemistry Department, Tshwane University of Technology, Private

Bag X680, Pretoria 0001, South Africa

and %Ndfa between wetland and upland P. pinnata plants.

These results suggest that, under limiting O2 conditions especially in wetlands, nodules make structural and functional adjustments to meet the O2 demands of N2-fixing bacteroids.
Keywords Inner cortex . Bacteria-infected central tissue .
Infected and uninfected interstitial cells . Bacteroids .
Morphometric analysis . Extracellular air space . Low pO2 .
15N . %Ndfa

Introduction
Nitrogen fixation in legume nodules has a high O2 demand for
oxidative phosphorylation at sites of bacteroid respiration but
low enough (30 nM) not to destroy nitrogenase (Appleby
1984,1992; Goodchild and Miller 1997). Maintaining this balance in O2 supply for nodule functioning has, in the past, been
the focus of many investigations (Atkins et al. 1990; Streeter
1995; Marino et al. 2006; Dalton et al. 2009). Nodules formed
in different artificially supplied pO2 revealed both anatomical
and physiological adaptations (Dakora 1990; Dakora and
Atkins 1989, 1990ab, 1991). Nodules developed in ambient
and supra-ambient pO2 maintain a steady flow of O2 to respiring bacteroids through the operation of a variable O2 diffusion
barrier (Witty et al. 1986, 1987; Dakora and Atkins 1989;
Bergersen and Turner 1990; Bergersen 1997; Loureiro et al.
1998) and the oxygenation of leghaemoglobin protein
(Wittenberg et al. 1972; VandenBosch and Newcomb 1988).
The adaptation of soybean nodules to different pO2 has been
reported to involve increased permeability to gases in those
grown at low pO2 and decreased permeability in those developed at ambient or supra-ambient pO2 (Dakora and Atkins
1991). The changes in ventilation characteristics of soybean
nodules formed in low pO2 included increased frequency of

S.A. Kanu, F.D. Dakora

lenticels, decreased nodule size, increased volume of cortex

relative to infected tissue, and changes in the size and frequency of extracellular voids in all tissue components (Dakora
1990; Dakora and Atkins 1991). Exposing cowpea nodules
to different O2 concentrations (180 %) for 28 days also resulted in marked anatomical changes in nodule size, lenticel
development, proportion of cortical vs. infected central tissue,
the frequency of infected to uninfected interstitial cells, the
volume of extracellular spaces in cortex and infected central
tissue, and the frequency of bacteroids (Dakora 1990; Dakora
and Atkins 1990ab). As a result of these structural changes,
the efficiency of bacteroid functioning increased from 0.10
0.02109 mol acetylene reduced per bacteroid in air-grown
nodules to 0.90.16109 mol acetylene reduced per bacteroid in 1 % O2-cultured cowpea nodules (Dakora 1990;
Dakora and Atkins 1990b).
In nature, some legumes are known to grow in habitats that
experience low pO2, often leading to adaptations that support
nodule functioning. For example, Viminaria growing in water
has been reported to form floating nodules, which form thick
lenticels for enhancing ventilation (Walker et al. 1983). A
number of studies (Subba-Rao et al. 1995; Rivas et al. 2009;
James et al. 1992, 2001) have also described nodulation, N2
fixation, and nodule structure in various aquatic and floodingtolerant legumes obtained from their natural environment. In
those legumes, the nodules formed on flooded stems and roots
developed thick layers of lenticels for facilitating O2 diffusion
to respiring bacteroids. In the Cape Fynbos of South Africa,
Psoralea species such as P. pinnata (L.) and P. aphylla (L.) are
also known to occur naturally on wetlands and well-drained
upland soils (see P. pinnata in Fig. 1a, b) and to effectively fix
N2 under both conditions. However, little is known about the
structure and functioning of P. pinnata nodules developed in
the two contrasting habitats. The aim of this study was to
determine nodule structure and functioning in P. pinnata
plants growing naturally in wetland or upland conditions in
the Cape Fynbos of South Africa.

Materials and methods

Plant sampling
Psoralea pinnata (L.) plants growing naturally in both wetland and well-drained upland soils in the Cape Fynbos of
South Africa (Fig. 1) were used for this study. With permission from staff of the Harold Porter Botanical Gardens, we
were allowed to destructively harvest only eight young P.
pinnata (L.) plants from the wetland part of the garden for this
study. The eight young plants were dug up with intact roots
and nodules from a P. pinnata plant population that was growing naturally in a stream (wetland) which ran several kilometers across the Botanical Gardens. Similarly, with permission

from the staff of the Kogelberg and Rooiles Nature Reserve,

we destructively harvested another set of eight young P.
pinnata plants along a hillside at nearby Kleinmond (upland
condition). In all instances, the nodulated plants were placed
in plastic buckets containing ice, transported to the laboratory
at the University of Cape Town, and separated into nodules,
stems, and branches plus leaves. Healthy, mature fully developed N2-fixing nodules were removed from roots and thoroughly washed with deionized water. The nodules were blotted dry, sectioned, and photographed (see Fig. 2).
The stems and branches plus leaves of the eight plants
harvested from the stream at Bettys Bay (wetland) and from
the hillside at Kleinmond (upland) were used to measure N2
fixation by the 15N natural abundance technique. Branches
with leaves were sampled from non-legume plant species
growing in the same wetland or upland environment as P.
pinnata and used as reference plants for estimating soil N
uptake by the legume. These non-legume reference plant species included Berzelia lanuginosa (L) Brongn, Othonna
quinquendenta Thunb, Pelargonium cucullatum (L.), and
Erica perspicua (L.) sampled from Kleinmond as well as
Passerina vulgaris (Meisn.), Cliffortia lifotia (Marloth),
Halleria lucida (L.), Kiggelaria africana (L.),
Anthospermum aethiopicum (L.), and P. cucullatum (L.) harvested from Bettys Bay (Table 1).

Plant processing for 15N analysis

The stems and branches plus leaves of each P. pinnata plant
sampled from the field were oven-dried (60 C) separately for
72 h, weighed, ground into fine powder (0.85 mm sieve), and
stored in tightly capped vials prior to 15N analysis. The nonlegume reference plant species were similarly processed for
15 14
N/ N isotopic analysis. The 15N and %N analysis were
done at the University of Cape Town using a Carlo Erba NA
1500 Elemental Analyzer coupled to a Finnigan MAT 252
isotope ratio mass spectrometer (Finnigan MAT GmbH,
Bremen, Germany) via Conflo II Open-Split Device.
The 15N/14N isotopic composition (15N) was measured as the difference in the number of atoms of 15N
to 14N in atmospheric (atm) N2 as (Mariotti et al. 1981;
Unkovich et al. 1994):

15 N0=00

15 N=14 Nsample 15 N=14 Natm

15 N=14 Natm

 1000

where the 15N value is the 15N natural abundance of

plant sample, expressed in a relative delta () notation,
which is the atom percent excess () deviation of the
sample from atmospheric N2 (0.3667 atom % 15N).

Structural adaptation of P. pinnata nodules to wetland conditions

Fig. 1 A population of Psoralea
pinnata (L.) plants growing in a a
wetland in the Harold Porter
Botanical Gardens in Bettys Bay
and on b a well-drained upland
soil on the Kleinmond mountains
in the Western Cape Province of
South Africa

Psoralea pinnata

Estimation of percent N derived from fixation

The percent N derived from fixation of atmospheric N2
(%Ndfa) was calculated from the 15N natural abundance of
the legume and the mean 15N of non-N2-fixing reference
plants from each site, as described by Shearer and Kohl
(1986) and Unkovich et al. (2008):



 .
15
%Ndfa 15 Nre f 15 Nleg
Nre f B  100

modified one-fourth-strength Hoaglands nutrient solution.

Three uninoculated jars containing P. pinnata seedlings were
included as control. After 14 months of growth, the plants
were harvested and separated into stems and branches plus
leaves, and each plant part was oven-dried (60 C) separately
for 72 h, weighed, and ground into fine powder (0.85 mm) for
15 14
N/ N isotopic analysis, as described above.

Nodule processing for microscopy

Where 15Nref is the mean 15N natural abundance of the
reference plants, 15Nleg the mean 15N natural abundance of
the legume, and B the 15N natural abundance of P. pinnata
plants that were solely dependent on N2 fixation for their N
nutrition.
The 15N of stems and branches plus leaves (b+l) of purely
symbiotic P. pinnata plants were measured and used to calculate whole-shoot B value as the average of the 15N of stems
and branches plus leaves weighted by the total N content of
stems and branches plus leaves (Robinson et al. 2000):
X

15 N whole shoot

15 Nstem  Nstem 15 N b l  N b l
X
Nstem Nb l

About 7 to 10 mature and effective nodules were removed from

each of the eight plants that were destructively harvested from
wetland and upland habitats. These fresh mature pinkish nodules were hand-sectioned to produce specimens with 0.6 to
1 mm thickness using razor blades. All the samples were immersed in 2.5 % glutaraldehyde in 0.1 M phosphate buffer (pH
7.4) for 24 h at room temperature. The samples were washed

where b+l=branches plus leaves, and whole shoot=stems+

branches with leaves.
Determination of B value for glasshouse-grown P. pinnata
plants
The B value is the 15N natural abundance of an inoculated test
legume raised on an N-free growth medium to ensure complete dependence on N2 fixation for its N nutrition. The B
value was determined for P. pinnata using seeds collected
from the Fynbos. Seedlings of P. pinnata were raised from
surface-sterilized seed in sterile Leonard jar assemblies
(Vincent 1970) and inoculated with broth culture of rootnodule bacteria isolated from P. pinnata plants. Three replicate
jars were used. The seedlings were grown with N-free

P
OC
MC
IC

Fig. 2 Medial section of a fresh root nodule from Psoralea pinnata (L.)
plant harvested from well-drained upland soil showing the periderm (P),
outer cortex (OC), middle cortex (MC), inner cortex (IC), and the
presence of leghaemoglobin (pink) in the infected central medulla (M)

S.A. Kanu, F.D. Dakora

Table 1 15N () values of branches plus leaves of reference plants
(n=46) used to estimate %Ndfa in Psoralea pinnata plants sampled
from wetland and upland sites
Site

Species

15N ()
Minimum

Maximum

Kleinmond

Berzelia lanuginosa (L) Brongn

Pelargonium cucullatum (L)

2.92
2.45

1.72
6.69

Bettys Bay

Othonna quinquendenta Thunb.

2.32

0.73

Erica perspicua (L)

Mean

8.61
5.362

7.98

Cliffortia lifotia (Marloth)

2.69

1.98

Passerina vulgaris (Meisn.)

Halleria lucida (L)

4.90
0.21

5.13
0.14

Kiggelaria africana (L)

3.12

1.99

Anthospermum aethiopicum (L)

7.91

6.99

Pelargonium cucullatum (L)

Mean

0.96
6.605

5.81

Transmission electron microscopy

twice with 0.1 M phosphate buffer (pH 7.4) for 5 min each time
and post-fixed in 1 % osmium tetraoxide (OsO4) for 2 h. Each
sample was washed with buffer for 5 min and washed again
twice with distilled water for 5 min each time and dehydrated
for 10 min each time using an increasing ethanol concentration
(30, 50, 70, 80, 90, and 95 %) and then twice in 100 % ethanol.
After that, the samples were washed twice with 100 % acetone
for a total of 30 min and then left overnight in a 3:1 mixture of
acetone and resin. The samples were next immersed in a 1:1
mixture of resin and acetone for 6 h, followed by a 3:1 mixture
for overnight. The samples were transferred into 100 % resin
for 6 h and into another 100 % resin for overnight and then
embedded in 100 % Spurrs resin (Spurr 1969) for polymerization at 60 C for 48 h.
Light microscopy
For light microscopy, the resin blocks were trimmed with
heavy-duty razor blades and the exposed embedded material
surface-smoothened with glass knives using ultramicrotome
before cutting thin sections (46 m) using a Reichert
Ultracut S ultramicrotome system (Reichert-Jung, Austria)
fitted with a diamond knife. About 24 thin sections per nodule were placed on slides and stained with 1 % toluidine blue

Table 2 Symbiotic performance

of P. pinnata in wetland and
upland conditions

or a 1:1 mixture of 1 % methylene blue and 1 % Azur II. The

sections were viewed, photographed using a camera fitted to
the microscope, and light micrographs developed. About 7 to
10 nodules per plant for eight plants per habitat were sectioned
for light and transmission electron microscopy.

For transmission electron microscopy (TEM), gold and

silver-colored, ultra-thin (0.51.5 m) sections were cut
using ultramicrotome, placed on 100-mesh, Formvarcoated copper grid, and stained with 2 % uranyl acetate
for 10 min, followed by Reynolds lead citrate for
5 min. Each grid was washed in 5 drops of distilled
water and blotted before transfer to a drop of lead citrate. The lead citrate was kept in a closed petri dish to
avoid reaction with excessive atmospheric CO2. The
grids were then washed 10 times in a series of water
drops and blotted dry using clean filter paper. The sections were viewed and photographed with a transmission electron microscope (JEM 1200 EXII (JEOL,
Japan) at different magnifications up to 10,000.
Photographic prints were made of the cell geometry
and cell arrangement in nodule tissue components, as
well as of bacteroids in symbiosomes within infected
cells and of extracellular air spaces in the inner cortex
and the central bacteria-infected tissue.
Morphometric analysis
Photographed nodule sections were analyzed using ImageJ
1.42i software. This software utilizes a java language to
acquire, analyze, and store measurements of areas, perimeters, Ferets diameter (length), and circularity from photographic prints. The software was used to estimate cell
parameters such as length, perimeter, and area from light
and electron micrographs. The perimeter is the length of
the outside boundary of the selected profile, and the
Ferets diameter is the longest distance between any two
points along the selected boundary. Additionally, the software calculates the profiles circularity (4*area/square
root of the perimeter, which is a shape descriptor that
gives a value for the selected profile or boundary. A value
of 1.0 indicates a perfectly round circle. As the value

Stems
Habitat
Wetland
Upland

%N
2.040.08b
2.790.02a

Branches+leaves
15N ()
2.960.08b
2.480.01a

%Ndfa
875.73a
985.42a

%N
1.230.22a
1.150.07a

15N ()
2.150.08b
1.250.15a

Values (meanS.E.) followed by similar letters in a column are not significantly different at p0.05

%Ndfa
792.49a
958.09a

Structural adaptation of P. pinnata nodules to wetland conditions

Table 3 Symbiotic performance
of P. pinnata plants sampled from
wetland and well-drained upland
soils

Whole shoot
Habitat

g DM

%N

15N ()

mg N

%Ndfa

Wetland
Upland

20.210.8a
11.923.2a

2.940.1a
3.660,4a

2.680.2a
2.100.0a

692.5156.7a
316.0965.46b

804.83a
890.9a

Values (meanS.E.) followed by similar letters in a column are not significantly different at p0.05. Whole
shoot=stems+branches+leaves. g DM=shoot dry matter weight in grams; mg N=N content in milligrams per
shoot

approaches zero, it indicates an increasingly elongated

shape. Cortical cells were assumed to be spherical and
their volumes calculated using the equation:
V

4 3
r
3

Where v=volume of profile and

r=radius of profile
Individual cortical cells were identified and counted on
light micrographs to estimate the number of cell layers in each
tissue component. Three sectors were randomly selected per
nodule section and the average number of cell layers estimated
for each cortical component per nodule. About 7 to 10 nodules
were used per plant for eight plants per habitat.

site. Thus, the combined mean 15N of all reference plants

sampled at Bettys Bay wetland was 6.605, while that of
the upland site at Kleinmond was 5.362 (Table 1). These
were thus the values that were used to estimate %Ndfa of P.
pinnata plants.
B value of purely symbiotic P. pinnata plants
The B value (or the 15N natural abundance of P. pinnata plants
relying solely on N2 fixation for their N nutrition) was measured for finely milled stems and branches plus leaves of
glasshouse-grown P. pinnata plants. The mean B values obtained for organs of P. pinnata were 2.43 for stems,
0.96 for branches plus leaves, and 1.69 for whole
shoots (i.e., stems+branches with leaves).

Statistical analysis
All measurements of cell area, length, volume, perimeter, and
circularity, as well as proportions of nodule components, and
symbiotic parameters (%N, whole plant 15N, %Ndfa, and Nfixed) were analyzed using one-way ANOVA and Statistica v.
10, StatSoft (USA). Where significant differences were found,
the means were separated using the Duncans multiple range
test at p0.05.

Symbiotic performance of P. pinnata plants grown

in wetland and upland conditions
Although stem nitrogen concentration was higher in P.
pinnata plants growing in well-drained upland soils, that of
branches plus leaves was similar under both wetland and upland conditions (Table 2). The 15N of stems and branches
plus leaves were consistently more negative and much lower
in wetland than upland P. pinnata plants (Table 2). However,

Results
N values of reference plants
15

The number of reference plants used in this study ranged from

four at Kleinmond to six at Bettys Bay. The 15N values of
branches plus leaves of the reference plant material were generally very negative, as found by Spriggs et al. (2003) for
many non-legume reference plant species in the Cape
Fynbos. As a result, the minimum and maximum 15N values
of reference plants used for calculating the combined mean
ranged from 8.61 to +6.69 at Kleinmond and from
7.91 to +5.81 at Bettys Bay (Table 1). To avoid any
bias in data handling, the 15N of all reference plant species
sampled (irrespective of variations) were used to calculate a
combined mean for estimating %Ndfa of P. pinnata at each

Table 4 Proportion of tissue components of the cortex and central

infected tissue of P. pinnata (L.) in medial sections of nodules
developed naturally in dry upland and wetland conditions in the Fynbos
% of sectional area of nodule occupied by each
tissue component in the two types of nodule
Tissue component

Upland

Wetland

Outer cortex

14.10.8aC

17.20.7aB

Middle cortex

25.82.3bB

54.54.8aA

Inner cortex

9.70.7aC

10.41.3aB

Central infected tissue 50.31.6aA

17.94.3bB

Values are expressed on area basis. Values (meanS.E.) followed by

dissimilar letters (in lower case) within a row for each tissue component
are significantly different at p<0.05. Values (meanS.E.) followed by
dissimilar letters (in upper case) within a column for upland or wetland
are significantly different at p<0.05. (n=7)

S.A. Kanu, F.D. Dakora

Table 5 Measurement of the perimeter, area, volume, and circularity of
cells in the inner cortex of root nodules of Psoralea pinnata plants grown
naturally in wetland and well-drained upland conditions
Measurement

Inner cortex
Upland

Wetland

Perimeter (m)
Area (m2)
Volume (m3)

167.4162.4b
3.230.2a
0.370.0b

260.5244.0a
4.591.3a
0.680.0a

Circularity

0.580.0b

0.710.0a

the percent N derived from fixation was greater in upland than

wetland plants due to a much lower reference plant 15N value
used to estimate %Ndfa in wetland P. pinnata plants. But more
importantly, the data from isotopic analysis of stems and
branches plus leaves showed that, even under wetland conditions, P. pinnata plants could meet between 79 and 87 % of
their N nutrition from symbiotic N2 fixation, while upland
plants obtained about 93 to 98 % of their N from symbiosis
(Table 2). Analysis of whole shoots (i.e., stems+branches
with leaves) further confirmed that P. pinnata from both wetland and upland conditions relied more on N2 fixation for their
N nutrition and respectively derived 80 and 88 % of their N
from atmospheric N2 fixation (Table 3). Shoot N concentration and whole-shoot 15N values were also similar for both
wetland and upland plants (Table 3).
Light microscopy and TEM studies
Fully developed P. pinnata nodules were of the determinate
desmodioid type (Sprent 2009). with 46 mm in diameter and
a spherical shape (Fig. 2). Microscopic examination of resinembedded sections of P. pinnata nodules from both wetland
and upland habitats showed the presence of six zones, namely
lenticels, periderm (outermost covering), outer cortex, middle
cortex, inner cortex (or nodule parenchyma), and the central
bacteria-infected medulla zone (Fig. 2). Measurements of nodule tissue components revealed marked differences in size
within habitat and between habitats (Table 4). For example,

Table 6 Measurement of the

perimeter and area of intercellular
air space in the inner cortex and
central infected tissue of root
nodules from P. pinnata (L.)
plants developed naturally in
wetland and upland conditions

nodules formed under upland conditions had very large central infected zone (50 % of whole nodule) compared with the
other cortical components (Table 4). By contrast, the central
infected tissue of wetland nodules was very small, forming
only 18 % of the total nodule sectional area when compared
to the cortex which formed 82 % of the nodule (Table 4).
A comparison of nodule tissue components from wetland
and upland P. pinnata plants also revealed marked size differences. Although the outer cortex was similar in size for both
wetland and upland nodules, the proportion of each wetland
nodule occupied by the middle cortex was 55 % compared to
only 26 % in upland nodules (Table 4). Conversely, the size of
the central medulla region was only 18 % in wetland nodules
compared to 50 % in upland nodules (Table 4). The size of the
inner cortex, though bigger in wetland nodules, was not statistically different from upland nodules.
Measurement of perimeter, area, volume, and circularity of cells in the inner cortex also revealed differences between nodules developed under wetland or upland conditions (Table 5). The inner cortical cells of
wetland nodules showed larger perimeters, in contrast
to those of upland nodules. As a result, they recorded
greater cellular volumes and higher circularity than their
counterparts in upland nodules (Table 5).
TEM micrographs of the central medulla region revealed the presence of uninfected interstitial cells and infected cells with symbiosomes containing bacteroids of
different shapes (rod-like, spherical, or oblate) in both upland and wetland nodules. Cell area, cell volume, and cell
circularity were measured for infected and uninfected cells
and found to be similar for each habitat (data not shown).
Unlike upland nodules, wetland nodules showed abundant
intercellular air spaces within the inner cortex and central
infected tissue. Measured on area basis, intercellular voids
were increased by 37 % in the inner cortex and 431 % in
the nodule medulla (Table 6). In fact, within the bacteriainfected zone, extracellular air spaces were abundant and
found at every three cell junction in wetland nodules
(Fig. 3). By contrast, these extracellular air spaces were
very few in nodules developed under well-drained upland

Intercellular air space

Measurement
Perimeter (m)

Inner cortex
Upland
69.622.1b (100)a

Area (m2)

0.490.0b (100)

Wetland

Central infected tissue

Upland
Wetland

124.618. 6a (179)
0.670.0a (137)

101.67.7b (100)
0.429.0b (100)

260.852.2a (257)
2.20.7a (531)

Values (meanS.E.) followed by dissimilar letters within a row for each measurement are significantly different
at p<0.05
a

Values in brackets represent % increase in area or perimeter of air space in nodules developed under wetland and
upland conditions

Structural adaptation of P. pinnata nodules to wetland conditions

Discussion
Morphology and internal organization of P. pinnata (L.)
nodules

Fig. 3 Light micrograph showing infected and uninfected cells and many
air spaces in the central infected tissue of a wetland nodule. These air
spaces were few in upland nodules. ICS intercellular space, I infected cell,
U uninfected interstitial cell

conditions (data not shown). Measurement of the perimeter

of extracellular air spaces in inner cortex and central infected tissue also revealed structural differences between
wetland and upland nodules. As shown in Table 6, the
perimeter of extracellular spaces in the inner cortex was
greater in wetland than upland nodules by 79 %. These
differences in the numbers (Fig. 3) and size of intercellular
air spaces were even more dramatic in the central medulla
region of the wetland nodules, with void perimeter increasing by 157 % and void area by 431 % when compared to upland nodules (Table 6). A count of bacteroids
in symbiosomes also showed significantly higher numbers
in upland than wetland nodules (Fig. 4).

The morphology and internal organization of N2-fixing P.

pinnata nodules were similar to those of cowpea and soybean,
which belong to the tribe Phaseoleae. Fully developed root
nodules of P. pinnata plants were of the desmodioid type
(Sprent 2009) with about 46 mm in diameter and a spherical
shape (see Fig. 2) without any fine rootlet subtending them on
the roots. As found with soybean (Pankhurst and Sprent
1975a, b) the surface of nodules from wetland P. pinnata
plants were distinctly covered with lenticels (data not shown),
anatomical features that were fewer on nodules of upland
plants. The function of lenticels is to permit the free movement
of gases into and out of the nodule and in so doing promote
internal ventilation of root nodules. Thus, their formation in
N2-fixing nodules tends to increase under conditions of low
pO2, which is a way of enhancing O2 supply for bacteroid
respiration, as would be expected under the wetland conditions in this study.
In addition to lenticels, microscopic examination of resinembedded nodule sections revealed five distinct nodule components. These included the periderm (which is derived from a
cork cambium, see Dakora and Atkins 1989) outer cortex,
middle cortex, inner cortex (or nodule parenchyma), and the
bacteria-infected central medulla region (see Fig. 2), as found
in determinate nodules of cowpea and soybean (Dakora and
Atkins 1989; Parsons and Day 1990). Also present were vascular bundles arranged at regular intervals in the middle cortex, similar to that reported for soybean (Parsons and Day
1990) and cowpea nodules (Dakora and Atkins 1989). The
presence of toluidine-blue dark-stained sclereid layer in the
cortex of P. pinnata nodules has also been observed in soybean nodules (Bergersen and Goodchild 1973). Additionally,
the central medulla region of the nodule consisted of bacteriainfected and uninfected interstitial cells, which showed lots of
dark-stained starch granules (Kaneko and Newcomb 1987;
Webb and Newcomb 1987).
Integrated structure-function relationship permitting
growth and N2 fixation of P. pinnata in Cape Fynbos
wetlands

Fig. 4 Average number of bacteroids per micrograph in infected cells of

wetland vs. upland P. pinnata nodules. Bars with different letters (a, b)
are significantly different at P0.05

Mass spectrometric analysis of plant samples generally

showed similar N concentrations in stems, branches plus
leaves, and whole shoots of young P. pinnata plants grown
in wetland and upland conditions in the Cape Fynbos
(Tables 2 and 3). Isotopic analysis of stems, branches plus
leaves, and whole shoots of wetland P. pinnata plants further
revealed that 87, 79, and 80 %, respectively, of organ N was
derived from symbiotic fixation when compared to 98, 93, and

S.A. Kanu, F.D. Dakora

88 % (in that order) in upland plants (Tables 2 and 3). This

clearly indicates that, despite the potentially low pO2 in the
wetland environment, P. pinnata plants were still highly dependent on N2 fixation for their N nutrition. The lack of differences in symbiotic parameters such as N concentration,
15N, and %Ndfa between whole shoots of wetland and upland P. pinnata plants (Table 3) suggests major anatomical
changes in root nodules, which increased O2 supply to N2fixing bacteroids. Light microscopy and TEM studies indeed
revealed marked differences in nodule ultrastructure and subcellular organization between P. pinnata plants developed in
wetlands and those on well-drained upland soils (Tables 4, 5,
and 6; Figs. 3 and 4).
Earlier studies have shown that root nodules developed
under low pO2 had massively abundant lenticels and a highly
reduced central medulla tissue with more intercellular voids
when compared to well-aerated nodules (Pankhurst and
Sprent 1975a,b; Dakora and Atkins 1990a,b, 1991). The anatomical adjustments observed in this study with wetland nodules are therefore consistent with those reported for cowpea
and soybean nodules exposed to artificial supply of subambient pO2 (Dakora and Atkins 1990a,b, 1991). Both the
organization of cells and intercellular voids in the inner cortex
bordering the central infected medulla region have been reported to be critical components for gaseous diffusion in functional nodules (Tjepkema and Yocum 1973; Dakora and
Atkins 1989; Layzell et al. 1993). Thus, the 37 % increase
in intercellular air space of the inner cortex, the 431 % increase
in extracellular air space of the nodule medulla, and the 181 %
decrease in the size of sectional area occupied by the medulla
of wetland nodules would no doubt increase gaseous permeability and enhance O2 supply to bacteroids, as found for
soybean and cowpea nodules grown artificially under subambient pO2 (Dakora and Atkins 1990a, b, 1991). More importantly, the predicted increase in O2 supply from anatomical
modifications (Fig. 3) to a significantly decreased number of
bacteroids in infected cells of wetland nodules (Fig. 4) could
markedly increase nodule functioning and symbiotic N yield
and thus explain the lack of differences in symbiotic performance between wetland and upland P. pinnata plants
(Tables 2 and 3). The net result of these marked anatomical
changes would be an increase in nodule internal ventilation,
greater bacteroid respiration, and increased N2 fixation for
plant growth.
The results of this study have shown clearly that, under
limiting O2 conditions, especially in wetlands, nodules can
evolve an integrated structure-function relationship to meet
the O2 demands of N2-fixing bacteroids. These structural
changes, when combined with the operation of a variable O2
diffusion barrier (Witty et al. 1986, 1987; Dakora and Atkins
1989; Loureiro et al. 1998) and the oxygenation of
leghaemoglobin (Appleby 1984, 1992) can substantially support nodule function and N2 fixation even under the most

limiting O2 conditions. Furthermore, it has been shown with

1 % O2-grown cowpea plants that, despite the nodules having
low level of leghaemoglobin, the amount of O2-binding protein per infected cell and/or per bacteroid was markedly increased by 2-fold and 5-fold, respectively, when compared to
those grown in air (Dakora et al. 1991) and in that order
resulted in 4.3 and 9.7-fold increase in nitrogenase activity
relative to air-grown nodules (Dakora and Atkins 1990a).
This suggests that nodules developed under O2-limiting conditions such as wetlands, can through anatomical, physiological, and biochemical mechanisms, maintain a high level of
symbiotic functioning to meet the N demands of the bacteroids and their host plant.
Interestingly, it has been shown elsewhere that Vigna
longifolia (L.), like P. pinnata, is also capable of growing in
seasonally flooded and drier areas of the Pantanal in Brazil
(Pott and Pott 1994) as well as in the wetland regions of South
and Central America (Hacker et al. 1996) and in the permanent wetlands of Puerto Rico (Dubey et al. 1972). While there
is currently little information on its symbiotic adaptation to
these low O2 conditions, it is likely that its nodules, like those
of P. pinnata, have evolved mechanisms that improve internal
ventilation and O2 acquisition and, in so doing, support bacteroid respiration and N2 fixation for optimal N nutrition of the
bacteria and their host plant. In the context of this study, nature
clearly has its own field experiments, which we duplicate in
laboratories through artificial exposure of nodulated roots to
varying pO2 (Dakora and Atkins 1990a,b; Parsons and Day
1990; Dakora and Atkins 1991; Dakora et al. 1991).
Acknowledgments This study was supported with grants from the
DST/NRF South African Research Chair in Agrochemurgy and Plant
Symbioses and the Tshwane University of Technology.
Compliance with ethical standards
Conflict of interest The authors declare that this study was conducted
in the absence of any commercial or financial relationships that could be
construed as a potential conflict of interest.

References
Appleby CA (1984) Leghaemoglobin and Rhizobium respiration. Annu
Rev Plant Physiol 35:443478
Appleby CA (1992) The origin and functions of haemoglobin in plants.
Sci Prog 76:365398
Atkins CA, Dakora FD, Storer PJ (1990) Effect of oxygen pressure on
synthesis and export of nitrogenous solutes by nodules of cowpea.
Planta 182:565571
Bergersen FJ (1997) Regulation of nitrogen fixation in infected cells of
leguminous root nodules in relation to O2 supply. Plant Soil 191:
189203
Bergersen FJ, Goodchild DJ (1973) Aeration pathways in soybean root
nodules. Aust J Biol Sci 26:729740

Structural adaptation of P. pinnata nodules to wetland conditions

Bergersen FJ, Turner GL (1990) Bacteroids from soybean root nodules:
respiration and N 2 -fixation inflow-chamber reactions with
oxyleghaemoglobin. Proc R Soc B 238:295
Dakora FD (1990) Studies of nitrogen fixation, nitrogen assimilation and
nodule development in grain legumeseffects of O2. PhD. Thesis,
University of Western Australia, Nedlands.
Dakora FD, Atkins CA (1989) Diffusion of oxygen in relation to structure
and function in legume root nodules. Aust J Physiol 16:131140
Dakora FD, Atkins CA (1990a) Morphological and structural adaptation
of nodules of cowpea to functioning under sub- and supra-ambient
oxygen pressure. Planta 173:572582
Dakora FD, Atkins CA (1990b) Effects of pO2 on growth and nodule
functioning of symbiotic cowpea (Vigna unguiculata L. Walp.).
Plant Physiol 93:948955
Dakora FD, Atkins CA (1991) Adaptation of nodulated soybean (Glycine
max L. Merr.) to growth in rhizospheres containing non-ambient
pO2. Plant Physiol 96:728736
Dakora FD, Appleby CA, Atkins CA (1991) Effects of pO2 on the formation and status of leghemoglobin in nodules of cowpea and soybean. Plant Physiol 95:723730
Dalton DA, Boniface C, Turner Z, Lindahl A, Kim HJ, Jelinek L,
Govindarajulu M, Finger RE, Taylor CG (2009) Physiological roles
of glutathione S-transferases in soybean root nodules 1[C][W][OA].
Plant Physiol 150:521530
Dubey HD, Woodbury R, Rodriguez RL (1972) New records of tropical
legume nodulation. Bot Gaz 133:3538
Goodchild DJ, Miller C (1997) Immunogold localization of hemoglobin
in Casuarina root nodules. Protoplasma 198:130134
Hacker JB, Glatzle A, Vanni R (1996) Paraguaya potential source of
new pasture legumes for subtropics. Trop Grasslands 30:273281
James EK, Sprent JI, Sutherland JM, Mcinroy SG, Minchin FR (1992)
The structure of nitrogen fixing root nodules on the aquatic
mimosoid legume Neptunia plena. Ann Bot 69:173180
James EK, Loureiro MF, Pott A, Pott VJ, Martins CM, Franco AA, Sprent
JI (2001) Flooding-tolerant legume symbioses from the Brazilian
Pantanal. New Phytol 150:723738
Kaneko Y, Newcomb EH (1987) Cytochemical localization of uricase
and catalase in developing root nodules of soybean. Protoplasma
140:112
Lazell DB, Diaz Del Castillo L, Hunt S, Kuzma SM, Van Cauwenberg O,
Oresnik I (1993) The regulation of oxygen and its role in regulating
nodule metabolism. In: Palcios R, Mora J, Newton EW (eds) New
horizons in nitrogen fixation. Kluwer, Dordrecht, pp 393398
Loureiro MF, James EK, Franco AA (1998) Nitrogen fixation by legumes
in flooded regions. In: Scarano FR, Franco AC (eds)
Ecophysiological strategies of xerophytic and amphibious plants in
the neotropics. Series Oecologia Brasiliensis, vol. IV. PPGE-UFRJ,
Brazil: Rio de Janeiro, pp 195233
Marino D, Gonzlez EM, Arrese-igor C (2006) Drought effects on carbon
and nitrogen metabolism of pea nodules can be mimicked by paraquat: evidence for the occurrence of two regulation pathways under
oxidative stresses. J Exp Bot 57:665673
Mariotti A, Germon GC, Hubert P, Kaiser P, Ltolle R, Tardieux A,
Tardieux P (1981) Experimental determination of nitrogen kinetic
isotope fraction: some principle; illustrations for the denitrification
and nitrification processes. Plant Soil 62:413430
Pankhurst CE, Sprent JI (1975a) Surface features of soybean root nodules. Protoplasma 85:8598
Pankhurst CE, Sprent JI (1975b) Effects of water stress in the respiratory
and N2-fixing activity of soybean root nodules. J Exp Bot 26:287
304

Parsons R, Day DA (1990) Mechanism of soybean nodule adaptation to

different oxygen pressures. Plant Cell Environ 13:501512
Pott A, Pott VJ (1994) Plantas do Pantanal. EMBRAPA-Centro de
Prequisa Agropecuaria do Pantanal, Corumb, MS, Brazil
Rivas R, Velazquez E, Willems A, Vizcano N, Subba-rao NS, Mateos
PF, Gillis M, Dazzo FB, Martnez-Molina E (2009) A new species
of Devosia that forms a unique nitrogen-fixing root-nodule symbiosis with the aquatic legume Neptunia natans (L.f.) Druce. Appl
Environ Microbiol 68:521752221
Robinson D, Handley LL, Scrimgeour CM, Gordon DC, Forster BP, Ellis
RP (2000) Using stable isotope natural abundances (15N and 13C)
to integrate the stress responses of wild barley (Hordeum
spontaneum C. Koch.) genotypes. J Exp Bot 51:4150
Shearer G, Kohl DH (1986) N2 fixation in field settings: estimates based
on natural 15N abundance. Aust J Plant Physiol 13:699756
Sprent JI (2009) Legume nodulation: a global perspective. John Wiley &
sons Ltd, United Kingdom, pp 133
Spriggs AC, Stock WD, Dakora FD (2003) Influence of mycorrhizal
associations on foliar 15N values of legume and non-legume
shrubs and trees in the fynbos of South Africa: implications for
estimating N2 fixation using the 15N natural abundance method,
Plant Soil 255:495502
Spurr AR (1969) A low-viscosity epoxy resin embedding medium for
electron microscopy. J Ultra Res 42:99107
Streeter JG (1995) A new model for the rapid effects of non-invasive
treatments on nitrogenase and respiratory activity in legume nodules. J Theor Biol 174:441452
Subba-Rao NS, Mateos PF, Baker D, Pankratz HS, Palma J, Dazzo FB,
Sprent JI (1995) The unique root-nodule symbiosis between
Rhizobium and aquatic legume, Neptunia natans (L. f.) Druce.
Planta 196:311320
Tjepkema JD, Yocum CS (1973) Respiration and oxygen transport in
soybean nodules. Planta 115:5972
Unkovich MJ, Pate JS, Sanford P, Amstrong EL (1994) Potential precision of the 15N natural abundance method in field estimates of
nitrogen fixation by crop and pasture legumes in South-west
Australia. Aust J Agric Res 45:119132
Unkovich MJ, Herridge D, Peoples M, Cadisch G, Boddey B, Giller K,
Alves B, Chalk P (2008) Measuring plant associated nitrogen fixation in agricultural systems. Australian Centre for International
Agricultural Research (ACIAR). Australian Government, Canberra
Vandenbosch KA, Newcomb EH (1988) The occurrence of
leghaemoglobin protein in the uninfected interstitial cells of soybean
root nodules. Planta 175:442451
Vincent JM (1970) A manual for the practical study of root nodule
bacteria. Oxford, Blackwell scientific
Walker BA, Pate JS, Kuo J (1983) Nitrogen fixation by nodulated roots of
Viminaria juncea (Schrad. & Wendl) Hoffmans (Fabaceae) when
submerged in water. Aust J Plant Physiol 10:409421
Webb MA, Newcomb EH (1987) Cellular compartmentation of ureide
biogenesis in root nodules of cowpea (Vigna unguiculata (L) Walp.).
Planta 172:162175
Wittenberg JB, Appleby CA, Wittenberg BA (1972) The kinetics of the
reactions of leghemoglobin with oxygen and carbon monoxide. J
Biol Chem 247:527531
Witty JF, Minchin FR, Skt L, Sheehy JE (1986) Nitrogen fixation and
O2 in legume root nodules. Oxf Surv Plant Mol Cell Biol 3:275314
Witty JF, Skt L, Revsbech NP (1987) Direct evidence for changes in the
resistance of legume root nodules to O2 diffusion. J Exp Bot 38:
11291140