1

Ministry of Agriculture & Rural Development

Collaboration for Agriculture & Rural Development 013/06VIE

Replacing fertiliser N with rhizobial
inoculants for legumes in Vietnam for
greater farm profitability and
environmental benefits
MS6: High Quality Inoculants Technical Report

September 2009

1
Table of Contents

1. Institute Information 2

Completion date (revised)
November 2009
Reporting period
December 2008 – September 2009

2. Contact Officer(s)
In Australia: Team Leader
Name:
Dr David Herridge
Telephone:
02 67631143
Position: Professor, Soil Productivity Fax: 02 67631222
Organisation
University of New England -
PIIC
Email:


In Australia: Administrative contact
Name:
Mr Graham Denney
Telephone:
02 63913219
Position:
Manager External Funding
Fax:
02 63913327
Organisation
Industry & Investment NSW
Email:
4. Executive Summary
Impact of two Australian strains in Vietnam on legume production and
productivity and comparative analysis between local and introduced strains
Part of this project was to evaluate elite international strains across the country and to
compare them with national strains. Included were local and imported strains from
Vietnamese institutes, from NifTAL (USA), ALIRU (Australia), DOA (Thailand), Korea and
Argentina. Several of these strains are currently used in commercial inoculants in Australia
such as CB1809 (soybean) and NC92 (groundnut). We conducted two experimental sets; the
first was in a potted field soil and the second in field trials.
In the potted field soil trial, there were 13 treatments for groundnut (11 groundnut strains, a

the aim that they may scale-up production and progressively take over supply as the
technology and markets are developed.

4
DakNong, Tay Ninh, Dong Thap, An Giang and Tra Vinh. There were at least 5 treatments in
each experiment:

1. Farmer
’
s practice without N fertiliser
2. Farmer’s practice with N fertiliser
3. Inoculation with Australian strains CB1809 (soybean) or NC92 (groundnut), -N
fertiliser
4. Inoculation with local strain: SL1 (for soybean) or GL1 (groundnut), -N fertiliser
5. Inoculation with local strain: SL2 (for soybean) or GL2 (groundnut), -N fertiliser

The Australian strains were the most effective in terms of nodulation, biomass yield and
grain yield. Compared with the uninoculated control, CB1809 and NC92 increased
nodulation of soybean and groundnut, respectively, by an average of 58%, biomass yield by
30% and grain yield by 29%. Compared to the local Vietnamese strains, CB1809 and NC92,
increased soybean and groundnut nodulation by an overall average of 22%. Biomass yields
were increased by an average of 10% and grains yields increased by an average of 13%.

Protocols for production of high quality inoculants including QA, packaging,
storage, distribution and on-farm application of inoculants
During the two years of the project, technology for inoculant production at the three institutes
(SFI, OPI and IAS) was developed. The principal aim was production of high quality of
inoculants containing >5 x 10
8
rhizobia/g and a maximum 1 x 10

standards for nitrogen fixing microbial fertilizers. However, it is very important to have
effective QA of legume (rhizobial) inoculants. A number of modifications to the Vietnam
National Standard for Nitrogen-Fixing Microbial Fertilizers (TCVN 6166-1996) were
justified to make it more relevant to rhizobial inoculants, based on production technology and
efficacy requirements. The new standards largely utilize the well-constructed and
comprehensive framework of the current standard. The proposed name of the standard is the
Vietnam National Standard for Legume Inoculants and contains details on the technical
requirements of the inoculants including labelling as well as methods of testing and
reporting.

Results of demonstration trials and effectiveness of demonstration trials in
improving farmer’s awareness of benefits
A total of 168 demonstration trials have now been conducted in 10 provinces. The
demonstration fields had two treatments: +inoculation with nil or very low amounts of
fertiliser N and –inoculation with farmer’s rate of fertiliser N. Results are summarised in
Appendix 3.

Generally, inoculation of soybean and groundnut increased the profit for farmers, on average
by 4.500.000VNĐ/ha. The size of the benefit varied across the different sites. The increase
was around 500.000VNĐ/ha at the demonstration field of groundnut at Bau Don, Tay Ninh
province, and as high as 14.200.000VNĐ/ha at Chau Thanh, Tra Vinh province. Similarly for
soybean, the profit from inoculation was as much as 11.640.000VNĐ at Duong Minh Chau,
Tay Ninh province. In Dong Thap province the benefit from inoculation was on average
4.900.000VNĐ/ha.

In Dong Thap province, at the Phu Huu vllage, Chau Thanh district where demonstrations
were conducted on a large area of land (61.5 ha) with the participation of 120 local farmers,
yields of soybean increased on average 12.5%, equal to 300 kg seed/ha. Farmers produced
higher incomes around 4.900.000 VNĐ/ha compared to their normal cultivation with N
fertilisers.

Research on legume inoculants in Vietnam has been done since the 1980
’
s at the Hanoi
University and SFI (VASI) in the North and, in the South, at Can Tho University (CTU), IAS
and OPI (now named IOOP). Generally, the objectives of the research were selection of
strains, small-scale production of inoculants and field trials evaluating efficacy of the
inoculants. Each institute focussed on target regions and particular legume crops, such as
CTU in the Mekong Delta with soybean, IAS in the Southern East Region with groundnut
and OPI in the Central Coast and Highlands with groundnut and soybean. Strains proposed
for inoculant production were not tested throughout the country and outcomes of associated
research on production technologies were not shared between the institutions. Thus, even
with a history of legume inoculant research and production in Vietnam, inoculants are
currently not available in the market and farmers are to a large extent unaware of their
potential benefits. Instead, farmers use expensive N fertilisers on their legume crops. Part of
this project was to evaluate elite international strains across the country and to compare them
with national strains. Included were local and imported strains from Vietnamese institutes,
from NifTAL (USA), ALIRU (Australia), DOA (Thailand), Korea and Argentina. Several of
these strains are currently used in commercial inoculants in Australia such as CB1809
(soybean) and NC92 (groundnut). We conducted two experimental sets; the first was in a
potted field soil and the second in field trials.

Methodology

Screening rhizobial strains in pots

The experimental design was a randomized complete block design with three blocks. There
were 13 treatments for groundnut (11 groundnut strains, a +N control without inoculation and
–N uninoculated control) and 18 treatments for soybean (17 soybean strains, a +N control
without inoculation and –N uninoculated control). Information of strains is given in the Table
1. Each strain was grown up in yeast mannitol broth (YMB) for 5–7 days to reach maximum

2
O - 20.63 mg/pot;
(NH
4
)
6
Mo
7
O
24
.7H
2
O - 0.81 mg/pot.

We planted 5 seeds/pot and removed 2 young plants after 7 days. The plants were harvested
at 30 days for soybean and at 45 days for groundnut. Numbers of nodules, dry weight of
nodules and dry weight of biomass was determined at harvest.

Table 1. Rhizobial strain information

No. Strain name Target crop Source
1 NC92 Groundnut Australia
2 Tal 179 Groundnut NifTAL
3 P088183 Groundnut Thailand
4 P03818 Groundnut Thailand
5 GL1 Groundnut OPI – local strain
6 GL2 Groundnut SFI – local strain
7 GL14 Groundnut OPI – local strain
8 LAC1 Groundnut SFI – local strain
9 P3 Groundnut OPI – local strain


6. Farmer
’
s practice without N fertiliser
7. Farmer’s practice with N fertiliser
8. Inoculation with CB1809 (for soybean) or NC92 (groundnut), -N fertiliser
9. Inoculation with local strain: SL1 (for soybean) or GL1 (groundnut), -N fertiliser
10. Inoculation with local strain: SL2 (for soybean) or GL2 (groundnut), -N fertiliser

Source of strains:

SL1: local strain (soybean) from Can Tho University
SL2: local strain (soybean) from SFI (VASI - from the national microbial strain program)
GL1: local strain (groundnut) from OPI
GL2: local strain (groundnut) from SFI (VASI - from the national microbial strain
program)
CB1809: Australian commercial inoculant strain (soybean) from ALIRU
NC92: Australian commercial inoculant strain (groundnut) from ALIRU

Measurements were: dry weight of nodules, biomass and grain yield. Plot size was at least 20
m
2
with 4 replications. A randomized complete block design was used. Depending on
growing areas, sowing date, land preparation, fertiliser inputs, date of sampling were
different. Details can be sent if required.

Peat inoculants were made by the three Vietnamese institutes (OPI, IAS and SFI) and in
some experiments Australian commercial inoculants were use as a positive control treatment.
Inoculant rates were 1–2 kg/ha for Vietnamese inoculants and 0.25 kg/ha for Australian
commercial inoculants. The inoculants were tested for quality by OPI before conducting

nodules/plant (mg)
Plant height (cm)
Dry biomass
(g/plant)
1. NC92 79 b 93 b 33 abcd 2.7 b
2. P08183 29 gh 35 fg 29 cd 1.7 ghi
3. CTP 44 e 52 d 36 ab 2.2 def
4. GL1 113 a 136 a 35 abc 3.8 a
5. P31 35 fg 42 ef 31 bcd 1.7 hij
6. GL2 77 b 91 b 35 ab 2.5 bc
7. P12 54 cd 66 c 35 ab 1.9 fgh
8. GL14 60 c 72 c 32 abcd 1.9 efg
9. LAC1 39 ef 47 de 38 a 1.6 ij
10. Tal 179 27 h 33 g 35 ab 1.5 jk
11. P03818 53 d 63 c 32 abcd 2.4 cd
Control 1 11 i 12 h 27 d 1.2 k
Control 2 5 i 6 h 32 abcd 2.2 de
CV% 10.6 11.4 12.7 8.6
Control 1: uninoculated, without N fertilizer
Control 2: uninoculated, plus N fertilizer (100ppm)
Source: OPI
Table 3. Nodulation and growth of inoculated soybean with different rhizobial strains
Rhizobial strains
Number of
nodules/plant
Dry weight of
nodules/plant (mg)
Plant height (cm)
Dry biomass
(g/plant)

In the first group, biomass amounts were 3 times higher than control treatments. The close
correlation between nodulation and biomass showed that nodules were very effective and
contributed to higher biomass yield (r
2
=0.86) and will likely play an important role in seed
yield increase. The 5 strains CB1809, U110, SL1, SL2 and CJ2 were the best strains for
soybean.

In another experiment conducted in a green house, CB1809 showed high effectiveness in
term of nodulation (Table 4). The strain produced more nodules compared to 3 other local
strains.

Table 4. Nodulation of 4 rhizobial strains for soybean in potted soil

Nodule number/plant No Treament
Total nodule
number
Nodule number
on main roots
Nodule number on
lateral roots
1 Control - - -
2 Inoculation with CB1809 50 25 25
3 Inoculation with SL1 32 13 19
4 Inoculation with SL2 39 13 26
5 Inoculation with SL3 32 11 22
Source: SFI
Field experiments
The total number of field experiments across the project was 36 in the 10 provinces. A
summary of project field experiments during 2007–09 and inoculation effects of CB1809 or

20
40
60
80
100
120
140
160
180
0 5 10 15 20 25 30 35 40
Field site
% Respons
eWith crop biomass, there were large responses (44–86%) to inoculation at 33% of the field
sites, moderate responses (17–40%) at 45% of the sites and small responses (5–19%) at the

12
remaining 22% of sites (Graph 2). Increases in grain yield from inoculation were smaller
than the increases in nodulation and biomass yield (Graph 3). There were large responses
(41–70%) at 20% of sites. Moderate responses (20–40%) were recorded at 62% of the sites
and small responses (4–18%) at the remaining 18% of sites. The overall average increases in
biomass yield and grain yield using the superior Australian strains were 30% and 29%,
respectively.

There were large differences in nodulation, biomass yield and grain yield responses amongst
the rhizobial strains. Australian commercial strains CB1809 (soybean) and NC92
(groundnut) were more effective than local Vietnamese strains at almost field sites
(Appendix 2). Data analysis shows that when the crops were inoculated with CB1809 or

18% sites, from 10–20% at 34% sites and from 1–10% at 48% sites. 13
Graph 4. Nodulation increases with CB1809 and NC92 compared to
local strains
0
10
20
30
40
50
60
70
80
0 5 10 15 20 25 30 35 40
Field sites
% increase
Local strain1
Local strain 2

Graph 5. Crop biomass increases with CB1809 and NC92 compared
to local strains
0
5
10
15
20
25
30

inoculants containing >5 x 10
8
rhizobia/g and a maximum 1 x 10
8
contaminants/g.

In the
following we present the current technologies for inoculant production in Vietnam. Some
details of the technologies are different between the collaborating institutes depending on
facilities and expertise. To some extent, the inoculant technologies have been adapted from
those used in countries with existing successful inoculant industries, eg Australia, US.
Strains for production
The project team has decided that CB1809 and NC92 will be used for inoculant production in
Vietnam as multi-field trials throughout the country showed these strains are the best for
soybean and groundnut. They increased nodule weight, crop biomass and grain yield
compared to local strains tested. In the future, more strain evaluation will likely be done to
try to develop even more effective inoculant strains.

Maintenance of strains and preparation of mother cultures

It is proposed for production of inoculants in Vietnam that mother cultures will be provided
annually by an independent quality control laboratory where the strains for production are
maintained in terms of purity, viability and effectiveness (nodulation and nitrogen fixing
ability). After receiving the mother cultures, producers (manufacturers) have to maintain
them for that year of production. 15
The first step after receiving the mother cultures is for the cultures to be transferred into
culture tubes containing YMA (yeast mannitol agar) medium (called sub-mother cultures), at

9
/ml broth. The broth in the flask can then be
aseptically poured into a fermentor.

Broth growth media: Medium for rhizobial growth consists of a carbon source, nitrogen
source and minerals. Most rhizobia can utilize pentoses, hexoses, disacharides,
polysaccharides and sugar alcohols even though the carbon utilization properties of rhizobia
vary. Generally, improvements in medium for inoculant production is based on the basic
medium, YMB (yeast mannitol broth). Large scale inoculant production requires cheap and
available ingredients, such as corn steep liquor and proteolyzed pea husks. Table 5 below
shows effects of different growth media on growth of 6 strains of rhizobia. The alternatives
tested by SFI were green bean extract, saccharose and glucose instead of expensive yeast
extract and mannitol. The number of rhizobial cells of the 6 strains yielded higher than
10
9
/ml in the proposed media SX1 and SX2, similar to YMB. The SX1 medium is the best
one in terms of efficacy and economic benefit.

The data are promising and need to be repeated. If rhizobial growth rates on media SX1 and
SX2 are again high, then they can be considered for inoculant production in Vietnam.

Duration of growth: The results showed that the numbers of rhizobial cells increased over
time and reached a maximum at the 7
th
day for all strains (Table 6). Counts were as high as 7
x 10
9
cells/ml.
NC92 2,8 x 10
9
4,8 x 10
9
5,6 x 10
9

GL1 5,6 x 10
9
2,8 x 10
9
1,9 x 10
9

GL2 5,6 x 10
9
7,2 x 10
9
5,2 x 10
9SX1(g/l): green bean extract 50 g; K
2
HPO
4
0.5 g; KH
2
PO
4

(Source: SFI)

Table 6. Number of rhizobial cells in broth cultures at different growth times

Rhizobial cells/ml
Growth time
(day)
CB1809 SL1 SL2 NC92 GL1 GL2
3 5,2 x 10
7
8,8 x 10
6
1,04 x 10
7
5,2 x 10
7
9,2 x 10
6
1,36 x 10
7

4 8,8 x 10
8
2,48 x 10
7
1,44 x 10
7
2,6 x 10
8
5,8 x 10

7 7,2 x 10
9
4,8 x 10
9
4,8 x 10
9
4,2 x 10
9
2,6 x 10
9
1,22 x 10
9

8 1,68 x 10
9
1,02 x 10
9
1,28 x 10
9
2,36 x 10
9
1,56 x 10
9
2,4 x 10
9

Source: SFI

pH effects:
Table 7. Effects of pH on the growth of rhizobial strains

5
CFU/ml)
++: nomarl development (10
6
–10
7
CFU/ml)
+++: good development (10
8
–10
9
CFU/ml
Source: SFI

Temperature effects:
Table 8. Effects of temperature on growth of rhizobia

Temperature
(ºC)
Growth of rhizobial strains
CB1809 SL1 SL2 NC92 GL1 GL2
25 ++ ++ ++ ++ ++ ++
30 +++ +++ +++ +++ +++ +++
37 + + + + + +
>45 - - - - - -
-: no development
+: weak development (10
4
–10
5

aeration. It is connected with the air outlet tube. Air will flow freely through both filters
while bubbling through the broth. The cotton wool in the filters needs to be packed uniformly
but loosely. Over packing the air inlet filter can cause resistance to incoming air and lead to
poor aeration. Over packing the outlet filter can lead to poor air escape and pressure build up
in the fermentor.

The pump is disengaged from the fermentor for sterilisation. Sterilise for 40 minutes at 1 atm
if the broth is 2 L and increase time by 10 minutes for each additional 1 L. After the
fermentor has cooled, remove the clamp from the air inlet tubing and connect the air supply
then check for proper aeration and for leaks in the system. The system is ready for
inoculation with the starter culture. Inoculation is conducted through the latex air inlet with a
sterilized syringe fitted with a needle. The air inlet tubing is surface sterilised with 70%
alcohol or 3% hydrogen peroxide about 2–3 cm above its connection to the glass tube. The

18
needle is inserted downwards into the tubing and the starter culture is injected. The culture
then is incubated at 30
o
C. Use of small fermentors would be appropriate for Vietnam,
particularly when combined with the broth dilution – solid state fermentation. With this, the
broth can be diluted 100 fold. Thus, a 2 L broth can be used to inoculate 10,000 packets (at
rate of 20 mL/packet).

Small steel fermentors are common in industry and are usually sterilized by autoclave. A
fermentor system for culturing rhizobia was designed by NifTAL and uses direct heating by
gas and cooling tubes for time-saving after sterilisation. The body of the NifTAL fermentor
is a pressure vessel with a 141 L total capacity. Working (broth) capacity is 20–100 L.
Details of this fermentor can be accessed through Prof. Nantakorn Boonkerd’s laboratory at
Suranaree University of Technology, Thailand.


Peat is the best researched and most frequently used as carrier materials for inoculant
production. A large number of studies have showed that rhizobia are protected and survive
well in peat. Tables 9 and 10 show physical and chemical analysis of well-researched peats.
These peat were used for commercial inoculant production in U.S.A and Australia. However,
physical and chemical analyses of a peat are only a partial assessement of its suitability as a
carrier. Only a test related to growth and survival of rhizobia can confirm its acceptability.

19
Apart from characteristics at the source such as salinity, clay, organic matter and
contamination with chemical residues, some unknown factors will affect suitability of peat
for use as an inoculant carrier. Peats from different sources should be tested after adjusting to
the same particle size distribution and moisture content (if possible). It is not possible to
judge the suitability of peat from the colour or texture. Table 10. Characteristics of Bendenoch Peat used for commercial inoculant production in
Australia

Characteristics Range Average
Organic matter (%) 28.8 – 75.4 64.3
Organic C (%) 16.4 – 42.1 36.1
Minerals (%) 10.0 – 16.0 12.1
Soluble salts (%) 0.09 – 1.50 0.87
Cl (%) 0.01 – 0.31 0.11
N (%) 0.89 – 2.30 1.83
K
2
O (%) 0.12 – 0.17 -
P
2

9

cfu/g moist peat.

20
Table 12. Number of rhizobia in different peat sources of the South

Treament Innitial 1 week 2 week 3 week 4 week
1. Australia 59 x 10
7
30 x 10
8
16 x 10
8
20 x 10
8
17 x 10
8
2. Komix 1 40 x 10
7
22 x 10
8
29 x 10
8
32 x 10
8
25 x 10

8
Treament 2 month 3 month 4 month 5 month 6 month
1. Australia 31 x 10
8
20 x 10
8
14 x 10
8
12 x 10
8
10 x 10
8
2. Komix 1 28 x 10
8
24 x 10
8
15 x 10
8
17 x 10
8
12 x 10
8
3. Komix 2 25 x 10
7
32 x 10
7
10 x 10
7
74 x 10
6

coconut coir dusts compared to peat control.
Table 13. Number of rhizobia in two peat sources of the North

Number of rhizobia/g moist peat inocualnts
Rhizobium
strain
Peat source
Initial 1 week 2 week 1 month 2 months 3 months
Sơn La 3,2 x 10
9
1,8 x 10
9
4,2 x 10
9
7,2 x 10
9
8,0 x 10
9
4,0 x 10
9
NC92
Thái Nguyên 2,4 x 10
9
1,3 x 10
9
3,6 x 10
9
7,6 x 10
9
4,0 x 10

1,8 x 10
9
7,5 x 10
9
3,4x 10
9
3,2 x 10
9
3,6 x 10
9
CB1809
Thái Nguyên 5,3 x 10
9
2,8 x 10
9
5,3 x 10
9
2,2 x 10
9
2,4 x 10
9
1,4 x 10
9
Sơn La 6,6 x 10
9
1,4 x 10
9
1,6 x 10
9
1,5 x 10

7

2 CB 1809 Peat 5.4 x 10
6

3 NC 92 Peat + worm casts 8.3 x 10
8

4 CB 1809 Peat + worm casts 5.4 x 10
8

5 NC 92 Peat + worm casts + coconut coir dust 2.5 x 10
9

6 CB 1809 Peat + worm casts + coconut coir dust 6.8 x 10
8

Source: IAS
1 NC92 Peat
1,2 x 10
8
2 NC92 Peat + molasses + “rare soil”
1,5 x 10
9

3 NC92 Peat + worm casts + coconut coir dust
3,8 x 10
9

Source: ISF

pH meter and gradually add lime until a pH of 6.5 has been reached
- Record the amount of lime needed to neutralize 10 g of the carrier
- Add the corresponding amount of lime to amount of the carrier. Mix well. 22
Water holding capacity: Water holding capacity of a carrier determines the amount of
liquid inoculum that can be added to the carrier. Carriers vary greatly in their water holding
capacity. Particle size, organic matter and clay content of peat will affect water holding
capacity and water potential. It is desirable to increase the water holding capacity of the peat
so that larger amounts of broth culture (and hence more cells) can be introduced to peat
before incubation.

The inherent moisture level of the carrier is determined. This is done conveniently on a
moisture balance but can use a drying oven if a moisture balance is not available. Weigh 10 g
of peat accurately on a foil or glass weighing disk and place it into the oven at 70
0
C in 48 hrs.
Check the end point of moisture loss.

Moisture content= [(W1 – W2) x 100%]/W2
W1: weight of carrier before drying
W2: weight of carrier after drying

Determination of the moisture holding capacity of the carrier. Weigh of 100 g of oven dried
carrier material into a 500 ml beaker. Add water with continue stirring until the carrier
appears to be saturated. Add additional water to produce a thin slurry. Transfer this slurry to
a pre-weighed measuring cylinder. Allow water to drain overnight, then weigh the measuring
cylinder with the contents. Give the moisture holding capacity on the dry weight basis of the
carrier. If 100 g of pre-dried carrier can hold 120 ml of water, its moisture holding capacity is

23
Table 15. pH and amount of broth cultures impregnated into moist sterilized peat
*Peat pH Amount of broth cultures
Australia 7 65
IAS 6.5 24
Nghe An 5 50
U Minh 6 60
DakNong 6 38
Komix 1 5 35
Komix 2 5 25

Source: OPI
Australian and IAS peat were already adjusted pH with lime
moist sterilized peat
*
: 20% moisture content

Calculations based on 70 g dry peat after adjusting to 20% moisture content for sterilisation

Table 16. Treatments for measuring optimum moisture content for legume inoculants

Moisture content (%) Liquid added (mL) Volume of broth (mL)
Volume of sterile water
(mL)
40 29.5 29.5 0
50 52.5 29.5 23
60 87.5 29.5 58

24
b)
Moisture to add to dry peat

35
120 100
x
x
=
+0.35(120 )
x
x
=
+

42 0.35
x
x
=
+0.35 42xx
−


Figure 7. Counting and confirmation of viable rhizobia from contaminated peats
10
-2
10
-3
10
-4
10

-5
10
-6
Carrier suspended in sterile water (10 g in 90 mL, 10
-1
dilution)
Spread 0.1 mL on the surface of duplicate
CRYMA plates and count colonies after growth
taking note of diltuions with contamination
Prepare dilution series to 10
-6
Inoculate 2 plants
from each of the
10
-5
and 10
-6
dilutions and
check for nodules
to confirm colonies
are rhizobia on the
corresponding
plates