August 2021
Appendix 4: Quantitative research
support to the Conservation Agriculture
Smallholder Farmer Innovation
Programme, KwaZulu-Natal.
Annual report
PROGRESS REPORT
Quantitative research support to the
Conservation Agriculture Smallholder
Farmer Innovation Programme, KwaZulu-
Natal
For period:
October 2020 to September 2021
Submitted to The Maize Trust
Compiled by:
E Kruger, M Dlamini, T Mathebula, N
Mdletshe, N Madondo, L Buthelezi, M
Malinga, N Mzobe and HJ Smith
Asset Research
In collaboration with:
Mahlathini Development Foundation
Contents
Contents ..................................................................................................................................................................................3
Background and project description..........................................................................................................................4
Approach and Methodology...........................................................................................................................................6
Key activities: October 2020-September 2021.....................................................................................................7
Budget .................................................................................................................................................................................8
Progress ..........................................................................................................................................................................10
Results achieved to date. ..............................................................................................................................................12
CA collaboratively managed trials (CMTs) ......................................................................................................12
Rainfall and runoff...........................................................................................................................................................13
Bulk density...................................................................................................................................................................16
Water productivity .....................................................................................................................................................19
Nematode indices .......................................................................................................................................................23
Soil Health ......................................................................................................................................................................24
Laboratory analysis ..............................................................................................................................................24
Sampling ....................................................................................................................................................................26
Different CA cropping options..........................................................................................................................34
Soil organic matter (SOM) ..................................................................................................................................36
Mycotoxin analysis .....................................................................................................................................................38
Progress summary ..........................................................................................................................................................41
Collaboratively managed trials (CMTs). ...........................................................................................................41
BERGVILLE ...............................................................................................................................................................41
MIDLANDS ................................................................................................................................................................49
Southern KZN ..........................................................................................................................................................57
Maize yield considerations .....................................................................................................................................58
Bergville maize yields ..........................................................................................................................................58
SKZN maize yields .................................................................................................................................................60
Midlands maize yields..........................................................................................................................................61
Issues and recommendations .....................................................................................................................................63
References ..........................................................................................................................................................................63
PHOTO CREDITS: Cover Page: Left: Strip crop planting byMr Cosmos Xaba (Madzikane, SKZN) and Right: Phumelele
Hlongwane (Ezibomvini, Bergville), strip crop planting of maize, beans and summer cover crops.
Background andproject description
Mahlathini Development Foundation (MDF) (2003-2021) is one of the few NGOs in South Africa
focussing on promoting collaborative pro-poor agricultural innovation. As such MDF is a
specialist NGO working in the fields of participatory research, training and implementation,
focussing on agroecological approaches, including conservation agriculture (CA).
The Maize Trust-funded CA Smallholder Farmer Innovation Programme (SFIP) in South Africa, as
conceptualized and implemented through MDF, has pioneered the use of agricultural innovation
systems as a methodological approach for the promotion of an appropriate smallholder CA
farming system, as well as awareness raising and adaptive research into specific elements of this
system (Kruger & Smith, 2019). This approach takes cognizance of the complexity of introducing
CA into a farming system,includingworking withsmallholder farmers as partners in the
knowledge co-creation process through on-farm research and experiential learning, as well as
embedding the process into the existing socio-political environments and economic value chains.
The overall goal of the CA SFIP is the mainstreaming of CA by grain farmers to ensure sustainable
use and management of natural resources while enhancing national and household food security
and income.
MDF has worked with smallholder farmers in CA learning groups, who implement CA as farmer
led trials and expand their implementation intotheir whole fields over time. Around 480
smallholders across 43 villages in KwaZulu-Natal (Bergville, SKZN and Midlands) have been
implementingCA for a period of between one to eight years, because of The Maize Trust SFIP
support. Over this periodseveral different indicators were designed and used to track the
progress of the participants.The table below outlines the progress, using a range of output and
impact indicators that have been monitored between 2014 and 2020.
Table Innovation system indicators for the CA-SFIP (2013-2020)
CA innovation system indicators for smallholder farmers in KZN; 2014 and 2020
Social agency indicators
Indicator
2014
Description/ comments
Participants
53
No of CA experimentation participants, from farmer registration lists across
all three areas
Learning groups
7
Count of no of village-based learning groups
Gender
89%
Percentage of women undertaking CA experimentation. Obtained from
farmer participation lists across all three areas
Local savings and loan
associations
0%
Percentage ofall learning group members involved inVSLAs (Village savings
and loan associations); fromsavings groups registers and learning group
membership lists
Innovation platforms
0
No of platforms set upthat include farmers and external stakeholders
Value chain indicators
Months of food
provisioning
No of participants, shown as a percentage who can provide enough maize
meal for their family for different month-based categories; from annual
review interviews for an average of 50 participants annually.
1 to 3
100%
4 to 6
0%
7 to 9
0%
10 to 12
0%
Local sale of crops
0%
No of participants, shown as a percentage who sell maize, beans,cowpeas
and sunflower produced, locally; from annual review interviews for an
average of 50 participants annually.
Saving for inputs
0%
No of VSLA members who used their savings and small loans for agricultural
inputs, shown as a percentage;from savings group records for 150
participants, averaged for a 3-year period
Farmer centres
0
No of farmer centres set up for sharing CA equipment, providing advice and
sale of agricultural inputs and produce between 2014 and 2020
Cooperatives
0
No of cooperatives registered for CA smallholders between 2014 and 2020
Co-financing of local
infrastructure
0
No of learning groups who took advantage of the matching grant funding to
finance local mills, threshers and water infrastructure or supplementary
irrigation
Mechanization
committees
0
No of committees set up within learning groups to manage the group owned
CA equipment, for use, hire and maintenance
Productivity indicators
Reduced labour in CA
plots
0%
No of participants, shown as a percentage who indicated a reduction of
labour throughout the cropping season; from annual review interviews for
an average of 50 participants annually, across all three areas.
Reduced weeding in CA
plots
0%
No of participants, shown as apercentage who indicated reduced weeding
in CA plots compared to conventionally cropped plots; from annual review
interviewsfor an average of 50 participants annually, across all three areas.
Use of CA planters
No ofparticipants, shown as apercentage using different CA planters
introducedthrough the programme; from planting and crop monitoring
forms, completed for between 50-200 participants annually, across all three
areas
Hand hoes
97%
Hand planters
0%
Animal-drawn planters
3%
Tractor-drawn planters
0%
Average Maize yield for
CA and conventional
plots (t/ha)
1,8 (CA),
2,2 (Conv)
Yield data measured and averaged for between 50 to 200 participants
annually across all three areas.
Crop rotation
0%
No of participants, shown as a percentage who practised intercropping of
maize and beans; from planting andcrop monitoring forms, completed for
between 50-200 participants annually, across all three areas
Intercropping - maize
and beans
0%
No of participants, shown as a percentage who practised intercropping of
maize and beans; from planting andcrop monitoring forms, completed for
between 50-200 participants annually, across all three areas.
Intercropping maize and
other legumes
0%
No of participants, shown as a percentage who practised intercropping of
maize and other legumes such as cowpeas and Dolichos beans; from planting
and crop monitoring forms, completed for between 50-200 participants
annually,across all three areas.
Winter cover crops
0%
No of participants, shown as a percentage who undertook planting of a
winter cover crop mixes (Saia oats, fodder rye, fodder radish, vetch, fodder
peas) from planting and crop monitoring forms, completed for between 50-
200 participants annually, across all three areas.
Cover crops: summer mix
0%
No of participants, shown as a percentage who undertook planting of a
summer cover crop mixes ((sunflower, millet, sun hemp, sorghum) from
planting and crop monitoring forms, completed for between 50-200
participants annually, across all three areas.
Seed saving
0%
No of participants, shown as a percentage who undertook seed saving of OPV
maize, legumes and cover crops; from planting and crop monitoring forms,
completed for between 50-200 participants annually, across all three areas.
Fodder: provisioning for
livestock: through cut and
carry, hay
0%
No of participants, shown as a percentage who cut and baled hay from their
CA plots and veld grass for winter feeding of livestock; from planting and
crop monitoring forms, completed for between 50-200participants
annually,across all three areas.
Reduced run-off in CA
plots
0%
No of participants, shown as apercentage who saw less run-off in their CA
plots when compared to their control plots; from planting and crop
monitoring forms, completed for between 50-200 participants annually,
across all three areas.
% Runoff for CA plots
13,5%
Runoff pans installed for 2-6 participants in each area and results averaged
for the croppingseason, for 2-4 years.
Percentage organic matter (%OM) per annum
Percentage organic matter measured and calculated for -5-8 participants
from eacharea, annually, after being averaged across all CA plotsfor each
participant.
Average annual sequestration of carbon is 0,3t/ha for CA plots(both trial
and control)
Midlands (2018 to 2020)
5,1%
SKZN (20182020)
5,7%
Bergville (20182020)
4,13%
Water Productivity
(kg/m3)
0,88
Water productivity is calculated as kg of crop produced per m3 of water. This
has been calculated for between 3 and 5 participants in each of the three
areas for maize grain and compared to conventionally tilled plots
For thepresent financial year, the intention isto focus on the adaptive research aspects of this
process, setting up quantitative collaborative-managed Trials (CMTs). In CMTs farmers and
researchers work together on problem definition, design, management and implementation of
trials as well as evaluation. These CMTs are considered the ‘Mother’”
1
trials and have been
implemented for 45 selected participants across the Bergville, SKZN and Midlands sites. They are
managed alongside the farmer led experimentation or “baby” trials, of which therehave been161
in this season.
Experimentation protocols chosen by the participants (researchers and farmers) in participatory
review and planning sessions are as follows:
1.10x10 blocks (1000m2): This denotes a 10-plot layout of100m2 plots, formultiple
cropping options (maize, legumes and cover crops), which is rotated on an annual basis.
2.Strip cropping (1000m2): Planting is done in 4m wide strips on contour, to provide for
soil and water conservation concerns and ease ofimplementation in larger fields, for
multiplecropping options (maize, legumes and cover crops), which is rotated on an
annual basis.
3.Short season maize:Planting of early maturing maize (white- PAN5A172 and yellow-
PAN5A190) alongside the normal varieties to test adaptability to climate variation, either
in block or strip cropping trials.
4.Fodder production: Planting of annual and perennial livestock fodder species (Teff, Tall
fescue, Lespedeza), forboth in situ grazing and baling, either in block or strip cropping
trials.
5.Cover crop seed production:250m2fenced plots are planted to 4-5 cover crop species
(sunflower, Sun hemp, Fodder Sorghum, Dolichos, turnips) with the specific intention of
keeping seed, for sharing and sale of cover crop seed to local CA farmers.
6.Poultry feed:Planting of 400-1000m2experimentation plots to poultry fodder species
(sunflower, Sun hemp, fodder sorghum), for harvesting of grain and preparation of
poultry rations for
7.Two-row planter:Introduction of two row minimum till planters for use by CA
participants to plant their larger CA trials and their own CA fields.
Approach and Methodology
36-48 CA participants from the already established CA learning groups volunteer to take part in
the collaboratively managed trials (CMTs),while other members of the groups also undertake
experimentation according to the 7 variations outlined above but are notsupported that closely
interms of quantitative measurements.
The process involves alearning workshop andpractical demonstration for each of the 7
variations, followed by intensive mentoring and support as each farmer implements their
experimentation in their own field.
1
Sieglinde S., Snapp S., DeDecker, J. and Davis, A. (2019). Farmer Participatory Research
Advances Sustainable Agriculture: Lessons from Michigan and Malawi. Agronomy Journal.
Volume 111, Issue 6. Pgs. 2681-2691.
In addition, with support from theLandCare division of the KZNDA, a larger group of farmers
undertake CA, but not thefarmer experimentation process.Most of thesefarmer participants
have been implementing CA for a number of years and they will be provided limited support and
mentoring to continue. The intention is to gather crop growth and yield data also from this larger
group, but to focus on the smaller groups for the quantitative data. This is to include thefollowing:
1. YIELD DATA:
Maize, yields under CA vs conventional cropping systems (100 participants)
Yield comparisons of different varieties of maize in the CA system(OPVs and Hybrids,
including short seasonvarieties); linked to planting dates, climate variability and pest and
disease resistance (36 participants)
Maize yields in CA rotation and intercropping systems (36 participants)
Biomass/dry matter yields for maize for water productivity assessments (9 participants)
Bean and cowpea yield in CA vs conventional tillage systems (50 participants)
Grain yields of cover crops planted (36 participants) and
Biomass /dry matter yields for livestock fodder (18 participants).
2. WATER RELATED DATA:
Local rainfall measurements in each village (x 8 villages)
Weather station and SAEON seasonal data; ET0, temperature, rainfall per site x 3 sites
Run-off (run-off pans installed in CA and conventional control plots and different CA
cropping options e.g. monocropping, intercropping, cover crops); 3-4 run-off pans per
participant (9 participants)
Bulk density assessments (9 participants)
Gravimetric water assessments; 3-4 treatments per participant (1 participant)
Water productivity assessments; 3-4 treatments per participant (9 participants)
3. SOIL RELATED DATA
Soil health analyses (Haney Soil Health Test (SHT), and
Nematode diagnostic indices) (24 participants, x treatments each, including veld, control
and one to two CA samples)
Key activities: October 2020-September 2021
CA farmer level experiments (1000m2block and strip cropping trials) were implemented by 192
participants across 22 villages in KZN (Bergville, ZKN and Midlands) in this season, with a further
189participants implementingCA under the KZN Landcare programme. The CMT plots have
been set upwith 9 participants in SKZN and Midlands respectively and with 27 participants in
the Bergville site.
Participantshave focused on experimentation in strip cropping, cover crops, different maize
varieties, use of 2 row tractor drawn planters,production of cover crop seed, and annual and
perennial livestock fodder species. Crop growth monitoring using an e-survey (Pendragon
platform) was conducted for 30 participantsfrom Bergville, 23 from SKZN and 20 from the
Midlands.
Rain gauges and run-off plots have been setup for 13 participants across as many villagesand
participant farmers have undertaken to keep ongoing records. The weather station in Ezibomvini
(Bergville)has provided accurate weather data for theBergvilleregion and SAEON has assisted
with further data from comparison with the farmer level records.
Gravimetric soil sampling has been undertaken for 2 participants; one from Bergville and SKZN
respectively. Thus far two sets of samples have been taken, dried and recorded.
Soil health sampleshave been taken for 24participants across 14villages and have been
submitted to Soil Health Support Centre (in the Western Cape) for Haney SHT analyses, as well as
nematode indices (through North West University, Potchefstroom) for comparison ofsoil health
results.
Stakeholder engagement and opendays havebeenseverely hampered by the continued COVID-
19 pandemic and movement restrictions. However, within these limitations,a cross visit was
conducted with 11 smallholder CA farmers from Ngongonini in Southern KwaZulu-Natal (SKZN)
to Bergville, specifically to view the collaboratively managed experimentation and processes and
a full-length article was writtenfor the March2021 edition of the African farmer magazine. Acase
study of our programme was submitted to the 8thWorld Congress on Conservation Agriculture
(8WCCA), for asession on the 23rdof Junefor experiences inAfrica related to sustainability and
mechanization.
Budget
The budget for this programme from October 2020-Spetember 2021 is R778 664,00. To date, end
June 2021expenditure has totalled R636 460,00. Thisleaves a monthly amount of R47 401,00 x
3 for finalization of the project.
Table 1: Budget summary for expenditure between October 2020 and September 2021
CA Adaptive research: Farmer Centred Innovation in CA. October
2020 September 2021; Maize trust
INVOICES
EXPENDITURE
Milestones/
Outputs
Key activities
OUTCOMES/
DELIVERABLES
Budgets
Oct
Nov
Dec
Jan
Feb
March
April
May
June
July
Aug
Sept
Actual exp
Pd by Asset
Research
TOTAL
EXPENDITURE
Farmer
experimentation
Bergville
Administration
and sundries
travel
accommodation,
admin,
accounting
R12 120,00
R11 085,25
R11
085,25
Farmer
centred
innovation
systems
farmer
experimentation,
researcher
managed
experimentation,
savings groups,
farmer centres…
R608 144,00
R 59 876,64
R63 030,20
R131 011,38
R67 182,45
R44 865,77
R37 674,22
R62 882,65
R63 437,98
R529
961,29
Analysis
Laboratory
costs, soil health
samples
R158 400,00
R0,00
R95 413,50
Monthly expenditure
R778 664,00
R 59 876,64
R 74 115,45
R0,00
R131 011,38
R67 182,45
R44 865,77
R37 674,22
R62 882,65
R63 437,98
R0,00
R0,00
R0,00
Overall expenditure
R636 460,04 (Actual end
June 2021)
R778 664,00
(Budget)
Progress
The tablebelow outlines activities related to objectives and key indicators for the period of
October 2020 to September 2021.
Table 2: Summary of progress (October 2020 September 2021) related to objectives and key activities.
Objectives
Key activities
Summary of progress
%
completion
Objective 1: To
assess the impact
of a range of CA
practices on
water, soil and
productivity
indicators, within
a smallholder
farmer level
experimentation
process
Key activity 1. Participatory
planning, design and layout of
experiments
160 participants across 20
villages plant CA Baby trials,
and Mother trails for 36 of
these participants
Key activity 2.Collection and
analysis of results
Quantitative measurements for
a minimum of 18 participants
(including run-off plots, soil
health analysis, bulk density
and water productivity
calculations, and yield
measurements (maize, cover
crops and fodder).
161 participants across 32 villages
have planted CA trials. 45 Mother
trials set up.
-Soil health samples and analysis
for 24 participants across 14
villages
-Rain gauges and run-off plots for
13 participants across 13 villages.
-Gravimetric water sampling; 2
participants across 2 villages
-Crop growth monitoring for 73
participants across 14 villages
-Water productivity samples (110)
for 13 participants from all three
areas for both grain and biomass
analysis
-Bulk density samples (27) from 13
participants across all three areas
taken and analysed.
-Mycotoxin samples (3-5 per
participant0 for 9 participants
across all three areas were taken
and sent to the ARC for analysis
-Yield measurements done for 96
participants across all three areas,
for maize, beans and cover crops
100%
100%
Objective 2: To
use results from
qualitative and
quantitative
results to outline
best bet options
for the
smallholder CA
farming system
and to provide
recommendations
for improvement
of the system
Key activity 1: Write up of
results
-Interim and annual reports
-Case studies (1 to 2 per site x
3 sites)
-Popular magazine and peer
reviewed articles (minimum of
1 each)
-Posting on appropriate
websites and online news-
letters and forums
Key activity 2: Presentation of
results
-Farmer review and planning
sessions (minimum 3,
maximum 6 sessions)
-Results presented to broader
farmer stakeholder forums for
networking and awareness
raising (1 to 3 events)
-Progress and annual reports
submitted.
-Short cases presented in this
report.
-African Farmer Magazine,” Farmer
Cheats” –January 2021 and full-
length feature February March
2021
-8WCCA case study for 23rd June
Africa session
www.mahlathini.orgwebsite
updated with latest documents.
Not done yet, planned for late
August-September
-Farmers cross visit from SKZN to
Bergville x 1
-Learner group presentations on
fodder production in other villages
and mycotoxin results x 3 in
Bergville
100%
100%
100%
0%
100%
A performance dashboard is indicated below. This provides a snapshot of performance according
to suggested numbers and outputs in the proposal.
Table 3: Performance dashboard; August 2021
Outputs
Proposed (March 2021)
Actual (Aug2021)
Number of areas of operation
3
3
No farmer level experiments
160
192
No of CMTs
36
45
No of local facilitators
14
VSLAs
18
Participatory monitoring and evaluation
process (farmer level)
60
73
Results achieved to date.
The table below outlines the villages, numbers of participants and experimentation processes for
the programme.
Table 4: Activities and numbers of farmers involved, per village for October 2020- August 2021
Area
No
Village name
(no of CMTs
or Mother
trials)
No of
participants
1000m2 trials
(10x10s)
400m2 trials
Strips
Fodder
species
Seed
Poultry
Two row
planter
Short season
maize
Actual
planted
(hectares)
Bergville
(12)
1
Eqeleni (3)
15
3
3
14
2
2
1,38
2
Ezibomvini (9)
16
11
4
2
3
3
1
7
1,46
3
Emafefetheni
8
2
8
0,88
4
Ndunwane
9
6
9
0,64
5
Stulwane (9)
38
26
12
10
5
4
6
10
4,08
6
Vimbukhalo (6)
24
3
5
13
2
2
3
10
1,8
7
Ngoba
2
2
0,2
8
Mhlwazini
11
4
7
0,38
9
Emabunzini
10
2
8
0,52
10
Emadakaneni
11
6
11
1,34
11
Emahlathini
9
1
9
0,94
12
Thamela
6
6
0,6
Midlands:
Ozwathini
13
Ozwathini (5)
38 (27)
12
15
8
5
1
6
4
3
2,72
Midlands:
Swayimane
(2)
14
Gobizembe (2)
25
3
15
7
4
1
6
2
8
1,6
15
Emayizekanye
(2)
35
7
28
7
8
1
2,32
SKZN (9)
16
Madzikane (3)
7
3
4
4
2
0,34
17
Springvalley (2)
13
13
0,52
18
Ofafa (1)
12
12
0,48
19
Ngongonini (2)
11
11
2
2
0,44
20
eMazabekweni
4
0,16
21
St Elois
9
0,36
22
Plainhill
9
9
0,36
23
Nkoneni (1)
17
11
1
1,7
24
Mission
11
11
0,44
365
73
184
119
29
8
17
22
44
25,66
CA collaboratively managed trials (CMTs)
The 7 experimentation protocols mentioned above (blocks, strips, short season maize, fodder
species, cover crop seed, poultry feed and use of the 2-row planter) were implemented within3
different layouts of CMTs:
1.1000m2(10x10m) plot (34replications),divided into 100m2and 10 blocks planted to
maize, beans, maize and bean intercrop, maize and cowpea intercrop, summer cover crop
mix (sunflower, sun hemp and fodder sorghum), Dolichos beans. The blocks are rotated
for those participants active for longer than 2 years, on an annual basis.
2.1000m2strip cropping plots(29 replications),with 2m wide strips of 4 rows of maize
alternated with aselection of legumes and cover crops, including for example beans,
cowpeas, Dolichos (Lablab), summer cover crops, teff, turnips, lucerneand perennial
fodder species such as Lespedeza, Pensacola and tall fescue. Short season maize was
included in the strip cropping trials as well.
3.250m2poultry feed or cover cropseed plots(25 replications),which arefenced and
planted to tramlines or strips of a selection of crops form which seed needs tobekept:
including Dolichos (Lablab), sunflower, sun hemp, millet, and fodder sorghum.
Plot layouts (2018-2020) were compiled for a total of 148 participants in Bergville, 37
participants in SKZN and 50 participants in Midlands.
Participants in the CMTs also undertake a control planting of at least 1000m2. Two types of
control were recorded:
1.CA-M control: For these plots, participants plant maize only in consecutive years, using
CA, but their own fertilizer and row spacing regimes. Average inter-row spacing for these
plots is between75cm to 90cm, so wider that the CA trial plots and participants use
fertilizer recommendations to determine fertilizer applications.
2.M-Conv control:For these plots, participants use conventional tillage andplant maize
only inconsecutive years. As farmershave moved across to using CAas their preferred
farming practise, theyhave stopped ploughing, meaningthat there are very few
conventional tillage controls. This season, only 2 participants in theMidlands ploughed.
All other control plots were CA-M controls.
Rainfall and runoff
This season run-off microplot pans have been installed for 13 participantsfrom all three areas.
The pans need to be installed after ploughing for the control plots, but prior to planting for both
the control and CA plots. This provided some logistical difficulties and some of the pans could
then only be installed in December 2020 and January 2021. It means that the data did not start in
October, as we were hoping.
Figure 1: Right and Fr
right: Installation of
run-off pans in control
and CA trial plots,
respectively.
The table below outlines the rain gauges and run-off pans installed.
Table 5: Rain gauges and runoff pans installed for 13 participants across Bergville, SKZN and Midlands for 2020-21
Village
Name and Surname
Rain
gauge
Run-
off
Plot
CA-
control
Ezibomvini
Phumelele
Hlongwane
Yes
Yes
Plot (2-SCC; 4-M+B; 6-M &9-Lablab)
Yes
Stulwane
Nelisiwe Msele
Yes
Yes
Plot (3-M+CP; 5-M; 7-M+B &10-SCC)
Yes
Vimbukhalo
Sibongile Mpulo
Yes
Yes
Plot 6
Yes
Eqeleni
Thulile Zikode
Yes
Yes
Plot (3-M+B; 5-Pumkin; 6-Lablab &9-SCC)
Yes
Ndunwana
Boniwe Hlatshwayo
Yes
Yes
Plot 1 (M+B)
Yes
Emabunzini
Valindaba Khumalo
Yes
No
No
No
Ofafa
Velephi Hadebe
Yes
Yes
Strip
Yes
Ngongonini
Mandla Mkhize
Yes
Yes
Plot 1 (M+B)
Yes
Springvalley
Letha Ngubo
Yes
Yes
Plot 2 (M+B)
Yes
Madzikane
Cosmos Xaba
Yes
Yes
Strip
Yes
Madzikane
Vakashile Gambu
Yes
Yes
Plot 2 (M+B)
Yes
Ozwathini
Doris Chamane
Yes
Yes
Strip
yes
Gobizembe
Rita Ngobese
Yes
Yes
Plot 5 (SCC)
yes
Mayizekanye
Dumazile Nxusa
Yes
Yes
Plot 1 (M+SCC)
Yes
Participant farmers have been provided with monitoring sheets to record rainfall events and run-
off for their CA trails and control plots. From previous experience it was thought to be an
acceptable strategy to obtain data. In addition, the participants who showed a lack of interest in
previous seasons were replaced with other participants and a few new farmers were brought on
board.
Onset of rain was again later than expected, being around themiddle to end of November.
Subsequent rainfall has been above normal, with very high levels of rainfall in February 2021.
Weather station data (Bergville) compared for the 2019/20 and 2020/21 growing season are
compared in the small table below.
Month
Rainfall 2019/20 (mm);
Weather Station SAEON
Didima
Rainfall 2020/21 (mm);
Weather station SAEON -
Didima
Rainfall 2020/21 (mm); rain
gauges average for Bergville
Oct
131,0
103,37
Nov
172,6
207,03
107,3
Dec
143,5
204,73
265,0
Jan
99,1
409,16
323,7
Feb
86,1
197,09
340,8
March
49,2
101,61
178,9
April
17,7
48,01
61,0
Totals
699,2
1 270,99
1276,7
From this small table the rainfall for the 2020/21 season has been much higher than the 2019/20
season. In addition, monthly averages have been consistently high between October- February,
while rainfall in the 2019/20 season dropped substantially after December 2019.
The cumulative rainfall between October2020/21as recorded by the farmersfrom the rain
gauges is very similar to the weather station data, even though the monthly averages vary
somewhat.Weather station data for the Midlands region, was taken from aweather station in
Swayimane, managed by the Umgeni Resilience Project (UKZN) in the area.
Averages have been calculated for monthly rainfall and runoff for each area.
Table 6: rainfall and runoff results for the 2020/21 cropping season for Bergville, Midlands and SKZN
Bergville (6 participants)
Month
Ave monthly rainfall
(mm)
Ave monthly runoff
CA plot (L)
Average monthly
runoff control plot (L)
Nov-20
107,3
2,8
Dec-20
265,0
15,1
16,6
Jan-21
323,7
30,2
62,0
Feb-21
162,7
15,9
22,7
Mar-21
178,9
4,5
5,1
Apr-21
61,0
3,9
4,0
Sum
1276,7
76,7
146,1
% rainfall conversion
6%
11%
SKZN (4 participants)
Month
Ave monthly rainfall
(mm)
Ave monthly runoff
CA plot (L)
Average monthly
runoff control plot (L)
Nov-20
83
Dec-20
148,3
10,2
14,0
Jan-21
141,8
0,5
1,2
Feb-21
132,8
1,6
3,3
Sum
505,8
12,3
18,4
% rainfall conversion
2%
4%
Midlands (3 participants)
Month
Ave monthly rainfall
(mm)
Ave monthly runoff
CA plot (L)
Average monthly
runoff control plot (L)
Nov-20
8,1
6,2
Dec-20
131,1
8,3
21,9
Jan-21
447,4
4,0
8,0
Feb-21
234,0
10,0
23,5
Mar-21
86,6
4,9
3,4
Apr-21
40,0
2,1
1,1
Sum
939,1
37,3
64,1
% rainfall conversion
4%
7%
Overall run-offin Bergville was substantially higher than the other two regions. The high
percentage clay in the soil in addition to high rainfall experienced this season has caused water
tables to rise, resulting in water saturation in the surface soil layers and thus higher levels of run-
off.
Figure 2: Right: An example
of a saturated soil profile in
Ezibomvini, Bergville in
mid- February 2021. The
small hole made by a soil
auger immediately fills up
with water. Far right: An
example of a field where
excessive rain caused
runoff and waterlogging in
a portion of the field.
For those participants where run-off between the CA trial microplots and Control microplots
were compared, the general trend noted in the last four seasons has remained the same. Runoff
is lower in the CA plots than inconventionally tilled control plots, or CA control plots where
monocropping has been practiced. Average runoff on CA trial plots is 4,3% of the rainfall and for
the control plots is 7,7%. Thus, runoff is cut by around 50% when conservation tillage and mixed
cropping is practiced.
Bulk density
Soil bulk density (BD), also known as dry bulk density, is theweight of dry soil (mass of solids)
divided by thetotal soil volume (Vsoil). Total soil volume is the combined volume of solids and
pores which may contain air (Vair) or water (Vwater), or both.
On cropland, long-term solutions to bulk density and soil compaction problems revolve around
decreasing soil disturbance and increasing soil organic matter. A system that uses cover crops,
crop residues and/or reduced tillage results inincreased soil organic matter, less disturbance and
reduced bulk density. Additionally, theuse of multi-crop systems involving plants with different
rooting depths can help break up compacted soil layers.
Samples were collectedfrom the top 5 cm of soil using sampling rings 7,2 cmin diameter using
the following procedure:
The ring was pushed (buried) into the ground using apiece of wood and a hammer (the
piece of wood was used to protect the ring).
A spade was used to dig the ring out of the soil.
Excess soil sticking out of the ring was cut using a knife to ensure the soil fitted perfectly
into the ring (making sure the volume was the same for all samples).
The soil samples (in thering) were wrapped with aluminium foil and transported to the
lab for analysis.
At the lab, samples were unwrapped, placed in aluminium dishes, weighed and assigned
codes, and put in an oven to dry at 100°C for 48 hours.
After 48 hours, sampleswere weighed and the masses were recorded for calculation of
dry mass.
The equation used to calculate the total soil volume is shown below.
   
Where, πis pi, r is radius and d is depth forthe ring, volume was calculated in cm3and mass of
the sample was measured in grams (g).
Average dry mass for all samples collected in the same plot was used to calculate the bulk density
and the same volume (based on the dimensions of the ring) was used. Equation 2 was used to
calculate the BD.


Bulk density samples were taken for 13 participants across all three areas, taking a range of CA
trial plot samples (M+B, M+CP, B, SCC, M) as well as samples from both conventional and CA
control plots planted to maize.
The table below outlines samples taken and the bulk density results.
Table 7: Bulk density samples and results for 13 participants from Bergville, SKZN and Midlands, June 2021
Area
Farmers name
Plot name/ number
Bulk Density (BD)
(g/cm3)
Bergville
Phumelele Hlongwane (PH)
Plot 1 (M+B)
1,182137
Bergville
Phumelele Hlongwane (PH)
Plot 4 (M+B)
1,173315
Bergville
Phumelele Hlongwane (PH)
Control CA
1,160083
Bergville
Nelisiwe Msele (NM)
Plot 3 (M+CP)
0,943946
Bergville
Nelisiwe Msele (NM)
Plot 7 (M+B)
1,182137
Bergville
Nelisiwe Msele (NM)
Plot 1 (M)
1,23948
Bergville
Nelisiwe Msele (NM)
Control CA
1,213014
Bergville
Sibongile Mpulo (SM)
Plot 1 (M+B)
1,305644
Bergville
Sibongile Mpulo (SM)
Plot 3 (M+B)
1,199781
Bergville
Sibongile Mpulo (SM)
Plot 6 (M+CP)
1,10274
Bergville
Sibongile Mpulo (SM)
Control Conv
1,257124
Bergville
Ntombakhe Zikode (NZ)
Strip 1 (M)
1,213014
Bergville
Ntombakhe Zikode (NZ)
Strip 7 (M)
1,323288
Bergville
Ntombakhe Zikode (NZ)
Strip 9 (M)
1,270357
Bergville
Ntombakhe Zikode (NZ)
Strip 3 (M+Pk)
1,14685
Bergville
Ntombakhe Zikode (NZ)
Control Conv
1,221836
Bergville
Boniwe Hlatswayo (BH)
Trial (M+B)
1,107151
Bergville
Boniwe Hlatswayo (BH)
Control CA
1,327699
SKZN
C Xaba (CX)
Plot 1 (M)
0,926302
SKZN
C Xaba (CX)
Plot 2 (B)
0,767507
SKZN
C Xaba (CX)
Plot 4 (SCC)
1,089507
SKZN
C Xaba (CX)
Control Conv
0,87337
SKZN
M Mkhize (MM)
M+B
1,093918
SKZN
M Mkhize (MM)
Control CA
0,966
SKZN
Thandiwe Hadebe (TH)
Trial (M+B)
1,182137
SKZN
Thandiwe Hadebe (TH)
Control CA
0,996877
SKZN
Letta Ngubo (LN)
Trial (M+B)
1,186548
SKZN
Letta Ngubo (LN)
Control CA
1,168904
Midlands
Mrs Xulu (MX)
Plot 2 (M)
1,138028
Midlands
Mrs Xulu (MX)
Control CA
1,076274
Midlands
Rita Ngobese(RN)
M (only)
1,032165
Midlands
Rita Ngobese(RN)
M+B
1,151261
Midlands
Babhekile Nene (BN)
Plot 1 (M)
1,05422
Midlands
Babhekile Nene (BN)
Control CA
1,085096
Midlands
Nomusa shandu (NS)
Plot 4 (Scc)
1,067452
Midlands
Nomusa shandu (NS)
Plot 1 (M)
1,279179
Midlands
Nomusa shandu (NS)
M+B
1,310055
Midlands
Nomusa shandu (NS)
Control Conv
1,279179
OVERALL AVERAGE
1,139305
Figure 3: Bulk density results for CA trial and control plots across Bergville, Midlands and SKZN, 2020/21
In general, bulk densities greater than 1,6 g/cm3 tend to restrict root growth (McKenzie, Jacquier,
Isbell, & and Brown, 2004). Sandy soils usually have higher bulk densities (1,31,7 g/cm3) than
fine silts and clays (1,11,6 g/cm3) because they have larger, butfewer pore spaces. In clay soils
with good soil structure, there is more pore space because the particles are very small, and many
small pore spaces fit between them. It has also been shown that minimum tillage practices can
increasebulk density without restricting aeration and water movement in thesoil, compared to
the same soils under tillage (Cavalieri, daSilva, Leão, Dexter, & Håkansson, 2009). Soils rich in
organic matter (e.g. peaty soils) can have densities of less than 0,5 g/cm3.
From Figure 3, the bulk density calculations for all cropping options in all three areas were within
the low to average range. Forboth Bergville and the Midlands, the bulk density results for the
Bergville MidlandsSKZN
Control CA1.23 1.08 1.04
Control Conv1.24 1.28 0.87
M+B 1.19 1.231.15
M+CP 1.02
M1.26 1.13 0.93
M+Pk 1.15
B0.77
SCC 1.07 1.09
0.00
0.20
0.40
0.60
0.80
1.00
1.20
1.40
g/cm3
Bulkdensity results for 13 participants across 3 areas, June 2021
conventionally tilled control maize plots were higher thanthe majority of the CA plots. The CA
plots planted to mixed crops with legumes and cover crops had lower bulk densities than the CA
maizeonly plots. For SKZN, most of the CA plot bulk densities were slightly higher than that for
conventionally tilled maize.
Bulk density sampling has been done in Bergville for 3 consecutive seasons.
Bergville ρb (g/cm3)
2018/19
2019/20
2020/21
CA multi cropped plots combined
1,26
1,29
1,12
CA control (M)
1,36
1,40
1,23
Conventional control (M)
1,30
-
1,24
Bulk density for conventionally tilled, CA control plots (M) and CA multi cropped plots do not vary
significantly, even though thegreatest reduction in bulk density has been for themulti cropped
CA plots. This could well belinked to the fact that increased soil organic matter in the fieldsin
Bergville has been challenging due to extreme weather variability and that ploughing is rare
enough that soil compaction is primarily due to the high density clay soils in the area, rather than
tillage. An averageρb of 1,2 g/cm3, 1,1 g/cm3and 1,0 g/cm3is used for thewater productivity
calculations for Bergville, the Midlands and SKZN respectively.
Water productivity
Crop water productivity (CWP)relates to the amount of yield per unit of water used. It is an
important measure of the impact of different practices on productivity in rain-fed agricultural
systems. Methods for improving CWP at field level include crop selection, planting methods,
minimum tillage, nutrient management and improved drainage, where appropriate. Average WP
for maize is 1,2-2,3 kg/m3 (FAO, 2003). In this research process, WP has been compared for different
crops and crop combinations under CA.
The main variables used in calculating water productivity (WP) are yields and volume of water
used to produce that yield. There are standard methods for working out the yield (e.g. putting the
harvested grain or biomass on a scale and weighing it, weighing a sample of maize cobs and
estimating yield using the plant population). The challenge is in determining the volume of water
used to produce the yield. There are a couple of methods (simple and more complicated) used in
determining the volume of water used.
To determine water productivity, parameters (temperature, relative humanity, solar radiation,
wind speed, wind direction to calculated ET0) are required and these parameters are gathered
from automatic weather stations. This information can be used to benchmark simpler methods
used in the field that farmers can be involved in. TheseET0 values are then multiplied by the crop
coefficient to find the actual evapotranspiration (Etc), whichis thevolumeof water used to
produce the yield.
To calculatetheET0, the equation below is used. The weather station calculates the reference
evaporation ET0 using the Penman Monteith equation shown below.
  
    
   
Where,
ET0 is reference evapotranspiration (mm/day)
Rn is net radiation at the crop surface (MJ m-2 day-1)
G is soil heat flux density (MJ m-2 day-1)
T is air temperature at 2 m height (°C)
u2 is wind speed at 2 m height (m s-1)
es is saturation vapour pressure (kPa)
D is slope vapour pressure curve (kPa °C-1)
g is psychrometric constant (kPa °C-1).
CWP for maize grain
Traditionally for intercropping, WP is calculated for the total grain in the plot, thus including
maize and beans/cowpea grain to get an overall view of the WP in the plot. The same is done for
the biomass. Maize grain yields are expected to be somewhat lower in intercropped plots than
in single cropped plots.
In this research process, WP has been calculated for maize only, both in single and intercropped
plots. The intention is to explore the WP of maize planted under different cropping options for
farmer participants who have been implementing CA for between 3 and 8 years.
Water productivity was calculated for nine participants KZN: 4 Bergville, 2 SKZN and 3 Midlands
for the 2019/20 cropping season. Waterproductivity for different CA cropping options was
calculated using the maize grain weight only, for mono cropped and inter-cropped plots.
The options were a M-CA control (consecutively mono-cropped maize); M-CA trial (single
cropped maize in a rotation system); M+CP-CA trial (maize and cowpea intercropped plot in a
rotation system), CA-M+B trial (maize and bean intercropped plot in arotation system)and
conventionally tilledmaizeas a control (Conv. Contr-M). Theaim was to ascertain whether the
different cropping options withinthe CAsystem provide fordifferent waterproductivity
outcomes. The results are shown in the figure below.
Figure 4: Water productivity for 9 participants across KZN 2019/20
NOTE: The WP results are for maize only, within each different cropping option
CA:M-ContrCA-MCA-M+BCA-M+CP Conv. Contr-
M
Total 0.80 1.111.21 1.43 0.36
0.00
0.50
1.00
1.50
2.00
Kg/m3
Water productivty for 9 participants across KZN 2019/20
For the2020/21 season the WP was calculated again for thedifferent CA cropping options, but
this time for 11 participants across KZN, as shown in the table below.
Table 8: WP calculated for 11 participants in KZN for different CA cropping options 2021/21
WP (kg/m3)
2020/21
Farmer name
M
M+B
M+CP
M-CA
Control
M-Conv
Control
Bergville
Boniwe Hlatshwayo (strip)
2,74
0,96
Nelisiwe Msele
2,57
2,32
2,10
1,01
Ntombakhe Zikode (Strip)
1,91
0,72
Phumelele Hlongwane
4,65
4,45
4,31
1,16
Sibongile Mpulo
3,27
3,04
2,12
1,44
SKZN
Cosmos Xaba (strip)
1,87
1,19
Letta Ngubo
3,52
2,08
Mandla Mkhize
2,15
1,00
Thandiwe Hadebe
2,55
0,92
Midlands
Mrs Xulu
1,92
0,20
0,98
Nomusa Shandu
0,44
1,01
0,53
Babhekile Nene (Jan'21)
1,87
0,85
0,63
Rita Ngobese
3,14
0,85
Overall averages
2,28
2,50
2,84
1,01
0,75
The results are summarized per cropping option in the graph shown below.
Figure 5: Water productivity for different CA cropping options in KZN 2021/21
What can be seen for the two figures for 2019/20 and 2021/21 is the following:
Cropping options
WP (kg/m3)
WP (kg/m3)
Ave WP (2
seasons)
2020/21
(n=11)
2019/20 (n=9)
CA Maize (M)
2,28
1,11
1,7
CA- Maize, bean intercrop (M+B)
2,50
1,21
1,9
CA- Maize cowpea intercrop (M+CP)
2,84
1,43
2,1
CA- Maize control (M-CA control)
1,1
0,8
1 ,0
MM+BM+CPM-CA ControlM-Conv
Control
Total 2.28 2.502.84 1.01 0.75
0.00
0.50
1.00
1.50
2.00
2.50
3.00
WP kg/m3
Water productivity (grain) for 11 participants in KZN 2020/21
Conventionally tilled maize (M-Conv Control)
0,75
0,36
0,6
Overall crop water productivity results were higher for all cropping options in 2020/21
than in 2019/20, indicating the different evapotranspiration effects of the two seasons.
For both seasons the WP for the conventionally tilled mono-cropped plots issubstantially
lower than for all the CA cropping options.
The WP for the CA control maize plots (which is planted to mono- cropped maize in every
season) is lower than the WP for the CA maize only rotation plots for both seasons. This
indicates a positive effect on WP dueto the rotation of maize with intercrops, legumes
and cover crops.
The WP for maize inthe legume (bean and cowpea)intercropped plots is substantially
higher than the maize only CA plots for both seasons. This indicates apositive effect on
WP for intercropping and closerspacing, which improves soil structure, organic matter
and reduces overall evaporation as a result of the soil cover.
The WP for maize in the M+CP intercropped CA plots is higher in both seasons than the
M+B intercropped CAplots, indicatinga more positive effect for cowpeas than beans on
WP. This is likely to bedirectly linked to the improved soil cover offered by cowpeas,
which grow with a much higher biomass than sugar beans as well as the nitrogen fixation
from cowpeas, also much higher than that of beans.
A small exploration of theeffect of inter-row spacing of maize is provided below. One of the
practices in the CA experimentation process is the use of closer spacing 50cm within and between
rows, than is the norm in these areas (75cm-1m between rows) and 50-75cm within rows.
If the CA M control samples forthe Bergville participants for example,are analysed for a 50cm
and 80 cm row spacing the results are as follows in the small table below.
Plot type
Spacing 50cm
Spacing 80cm
WP (kg/m3)
M CA-control
2,00
1,72
M
2,61
M+B
2,63
M+CP
2,84
This indicates a definite effect of row spacing on WP, indicating a positive effect on WP for closer
spacing options in Bergville.
Water productivity for CA maize grownas an intercrop with beans or cowpeas is higher
than single cropped CA maizeand water productivity forCA plots ishigher than
conventionallytilled plots. Despite annual differences in water productivity, these trends
remained the same across two seasons for all three areas within KZN. The close spacing
usedin the CA trial plots provides extra WP benefits when compared to the ‘normal’
spacing used in these villages.
WP for biomass
The CWP was also calculated for the above-ground maize plant material (stalks and leaves).
Figure 6: WP for grain and biomass for maize, for 11 participants in Bergville 2020/21
Again, there is avery clear trend of much improved WP for maize in theCA trial system when
compared to the CA maize control and conventionally tilled maize. The expected effect of lower
maize biomass yields in intercropped plots, compared to single cropped plots is also evident. The
significance of this result lies in the improved maize biomass WP for the rotation and multi-
cropped plots under CA, providing a clear indication of the benefit of this system in improving
crop water use over time.
Nematode indices
Nematode indices were reported upon in detail in the 6-monthly report, submitted in March
2021. In summary the Nematode indices indicated that fields where CA has been implemented
for 5-6 years show atrend of greater nutrient availability, greater ecosystem structure and
stability (with the nematodes populations as the proxy for this) and amove from bacteria to
fungal dominated energy pathways.
What the analysis showed, is that nematode populations or indices, as an indication for the soil
ecosystem or soil biology, changes (as a trend) slowly over time, with the first ‘visible’ signs of
change being a shift to a more fungal dominatedecosystem followed in subsequent yearsby a
greater structure and maturity of the nematode populations.
The analysis also showed that there were no significant differences in nematode indices between
the different CA treatments, such a M, M+B and SCC plots. In summary, all the CA treatments,
including theCA Maize control need to be considered important for improvement of soil
ecosystem health and each has a slightly different effect and impact on the soil ecosystem. SCC
plots for example provide for the presence of the highest number of population types and
numbers of nematodes as well as linkages between soil food webs but pushes the system
temporarily into a more bacterially dominated pathways, due tohigher enrichment values. M+B
intercropped plots have the highest channel index figures and the lowest enrichment values of
the treatments, thus favouring fungal dominated soil ecosystem development.
MM+BM+CP M-CA
Control
M-Conv
Control
Average of WP grain (kg/m3)2.28 2.50 2.84 1.01 0.75
Average of WP biomass (kg/m3)2.56 2.29 1.28 0.82 0.88
0.00
0.50
1.00
1.50
2.00
2.50
3.00
WP kg/m3
Water productivity (grain and biomassof maize) for 11 participants in
KZN 2020/21
Soil Health
The intention is to compare the soil health characteristics for several cropping options within the
CA trials, with conventionally tilled mono-cropped control plots, over time.
The soil health tests (as analysed by Soil Health Support Centre in the Western Cape and Ward
Laboratories in the USA) provide insight into microbial respiration and populations in the soil,
organic and inorganicfractions of the main nutrients N, P and K, and assessment of organic carbon
and percentage soil organic matter (% SOM). An overall soil health score (SH) is also provided for
each sample.
Method
SAMPLING
Sampling is done at thesame time every year, during September, after harvest and prior to the
start of seasonal rain, according to international conventions (Stolbovoy, et al., 2007).
CA plots:10 m x 10mplots are marked, and 10 cm depth cores are taken (with a soil
auger), taking 20 samples along a zigzag pattern across the plot. These are combined,
thoroughly mixed and then 500 g isplaced in a plastic bag and sealed. These bags are kept
in a cool, dark place until delivery to thesoil health analysis laboratory usually within
four to six weeks of taking the sample.
Control plots: 20 samples are taken in a zigzag pattern across the dimension of the control
plot; these vary from one participant to the nextand are otherwise treated in the same
manner as the CA plot samples above.
Veld samples: This changedafter the firsttwo seasons to reduce potentialvariability in
the samples. Apatch ofundisturbed veld, as close as possible to theparticipant’s cropping
field was chosen, to also have the same basicvisual characteristics as the field in question.
Four subsamples were taken at 10 cm depth at the four compasspositions adjacent to the
cropping field.
Samples were air dried and stored for a period of two to four weeks at room temperature
(20-24°C), prior to analysis.
Laboratory analysis
Laboratory analysis was undertaken bySoil Health Support Centre
(https://www.soilhealthlab.co.za/), linked to WARD Laboratories (https://www.wardlab.com/)
in the USA. Each soil sample received in the lab is dried at 50°C for 24 hours and ground to pass
a 4,75 mm sieve. The dried and ground samplesare scooped, with the weight recorded using a
Sartorius Practum 2102-1S, into two 50ml centrifuge tubes (4 g each) and one 50 ml plastic
beaker (40 g)that is perforated and has a Whatman GF/D glass microfibrefilter to allow water
infiltration. The two 4 g samples are extracted with 40 ml of DI water and 40 ml of H3A
respectively, for a 10:1 dilution factor. The samples are shaken forten minutes, centrifuged for
five minutes, and filtered through Whatman 2V filter paper. The water and H3A extracts are
analysed on a Seal Analytical rapid flow analyser for NO3-N, NH4-N, and PO4-P. The water extract
is also analysed on an Elementar TOC select C:N analyser for water-extractable organic C and total
N. The H3Aextract is also analysed on an Agilent MP-4200 microwaveplasma forAl, Fe, P, Ca,
and K.
The 40 g soilsample is analysed for CO2-C ppm after a 24-hour incubation at 25°C. Initially, the
sampleis wetted through capillary action by adding 18 ml of DI water to an 8 oz. glass jar (ball
jar with a convex bottom), placed in the jar and then capped. Solvita paddles can be placed in the
jar at this time and analysed after 24 hours with a Solvita digital reader. Alternatively, weuse a
system that we call HT-1, where, at the end of 24-hour incubation, the CO2in the jar can be pulled
through a LiCor 840A IRGA, which is a non-dispersive infrared (NDIR) gas analyser based upon a
single-path, dual wavelength infrared detection system.
SOM% is a gravimetric expression of the organic material fraction lost from combustion at 360°C
for three hours and is also termed the loss in ignition calculation method (LOI%).
Soil health test parameters
2
The method uses nature’s biology and chemistry by: (1) using a soil microbial activity indicator;
(2) a soil water extract (nature’s solvent); and (3) the H3A extractant, which mimics the
production oforganic acids by living plant roots to temporarily change the soil pH, thereby
increasing nutrient availability.
These analyses are benchmarked against natural veldfor each participant, due to high local
variation in soil health properties, and measured at different times. The veld scores provide for
high benchmarks against which to compare the cropping practices.
Soil Respiration one-day CO2-C:This result is one of the most important numbers in this soil
test procedure. This number (in ppm) is the amount of CO2-C released in 24 hours from soil
microbes after soil has been dried and rewetted (as occurs naturally in the field). This is a
measure of the microbial biomass in the soiland is related to soil fertility andthe potential for
microbial activity. In most cases, the higher the number the more fertile the soil.
Microbes exist in soil in great abundance. They are highlyadaptable to their environment and
their composition, adaptability, and structure are a result of the environment they inhabit. They
have adapted to the temperature, moisture levels, soil structure, crop and management inputs, as
well as soil nutrient content. Since soil microbes are highly adaptive and are driven by their need
to reproduce and by their need to acquire C, N, and P in a ratio of 100: 10: 1 (C:N:P), it is safe to
assume that soil microbes are a dependable indicator of soil health. Carbon is the driver of the
soil nutrient-microbial recycling system.
Water extractable organic C(WEOC): Consists of sugars from root exudates, plus organic
matter degradation. This number (inppm) is theamount of organic C extracted from the soil with
water. This C pool is roughly 80 times smaller than the total soil organic Cpool (% Organic Matter)
and reflects the energy source that feeds soil microbes. A soil with 3% soil organic matter when
measured with the same method(combustion) at a0-10cm sampling depth produces a 20 000
ppm C concentration. When thewater extract from thesame soil is analysed, the number typically
ranges from 100-300ppm C. Thewater extractable organic Creflects the quality of theC in the
soil and is highly related to the microbial activity. On the other hand, percentage SOM is about the
2
Haney/Soil Health Test Information Rev. 1.0 (2019). Lance Gunderson, Ward Laboratories Inc.
quantity of organic C. In other words, soil organic matter is the house that microbes live in, but
what is being measured is the food they eat (WEOC and WEON).
If this value is low, itwill reflect intheC02evolution, whichwill also be low. So, less organic carbon
means less respiration from microorganisms, but again thisrelationship is unlikely to be linear.
The Microbially Active Carbon (MAC= WEOC/ppm CO2) content is an expression of this
relationship. If the percentage MAC is low, it means that nutrient cycling will also be low. One
needs a %MAC of at least 20% for efficient nutrient cycling.
Water extractable organic N (WEON):Consists of Atmospheric N2sequestration from free
living N fixers, plus organic matter degradation. This number is the amount of the total water-
extractableN minus the inorganic N (NH4-N + NO3-N). This N pool is highly related to thewater
extractableorganic C pool and will be easily broken down by soil microbes and released to the
soil in inorganic N forms that are readily plant available.
Organic C:Organic N:This number is the ratio of organic C from the water extract to the amount
of organic N inthe water extract. This C:N ratio is acritical component of the nutrient cycle. Soil
organic C and soil organic N are highly related to eachother as well as thewater extractable
organic C andorganic N pools. Therefore, we use the organic C:N ratio of the water extract since
this is the ratio the soil microbes have readily available to them and is a more sensitive indicator
than the soil C:N ratio. A soil C:N ratio above 20:1 generally indicates that no net N and P
mineralisation will occur. As the ratio decreases, more N and P are released to the soil solution
whichcan be taken up by growing plants. This same mechanism is applied to the water extract.
The lowerthis ratio is, the more organisms are active and the more available the food is to the
plants. Good C:N ratios for plant growth are<15:1. The most ideal values for this ratio are
between 8:1 and 15:1.
Soil Health Calculation:This number is calculated as one-dayCO2-C/10 plus WEOC/50 plus
WEON/10 to include aweighted contribution of water-extractable organicC and organic N. It
represents the overall health of thesoil system. It combines five independent measurements of
the soil’s biological properties. The calculation looksat the balance of soil C and N and their
relationship to microbial activity. This soil health calculation number can vary from 0 to more
than 50. This number should be above seven and increase over time.
Sampling
Sampling has been done for each of the three areas for between 3 and 5 seasons. As the SH results
are not comparable across areas, they will be tackled per area in the discussions below.
Area
Village
2020/21
2019/20
2018/19
Bergville
Stulwane
4
3
2
(5-8 years)
Ezibomvini
3
3
2
Ndunwana
1
2
2
Eqeleni
2
-
2
Vimbukhalo
2
-
-
Mhlwazini
-
-
2
SKZN
Madzikane
2
2
2
(3-5 years)
Ngongonini
1
1
1
Spring Valley
1
1
1
Ofafa
1
1
1
Midlands
Gobizembe
1
1
1
(1-3 years)
Mayizekanye
2
3
3
Ozwathini
3
-
-
For the2020-21season soil health samples have been collected from 24 participants across 12
villages. Ozwathini (Midlands) has now been included as the participants there have been
practicing CA for a period of between 1-3 years.
The table below outlines samples taken across the three areas.
Table 9: Soil samples taken for 24 participants across all three areas.in September 2020
Name
Surname
Trial plot
Control plot
Veld
No. of soil
samples
Ezibomvini
Phumelele
Hlongwane
M, M+B, SCC, Dolichos
(Plots 2,4,6,9)
Maize (hand hoe)
yes
6
Zodwa
Zikode
M, M+B, SCC,
Maize (hand hoe)
yes
5
Nombono
Dladla
M, M+B, SCC
Maize (hand hoe)
yes
5
Ndunwana
Boniwe
Hlatshwayo
M+B
Maize (hand hoe)
yes
3
Stulwane
Dlezakhe
Hlongwane
M, M+B, SCC
Maize (hand hoe)
yes
5
Khulekani
Dladla
M, M+B, SCC
Maize (hand hoe)
4
Nelisiwe
Msele
B, M, SCC
Maize (hand hoe)
yes
5
Cuphile
Buthelezi
M, M+B, SCC
Maize (hand hoe)
yes
4
Matolozana
Gumbi
M, M+B, SCC
Maize (hand hoe)
4
Eqeleni
Ntombakhe
Zikode
M, M+B, SCC
Maize (hand hoe)
yes
5
Thulile
Zikode
M, M+B, SCC
Maize (hand hoe)
4
Vimbukhalo/Emazimbeni
Sibongile
Pulo
M, M+B
Maize (hand hoe)
yes
3
Zweni
Ndaba
M, M+B
Maize (hand hoe)
4
Springvalley
Letta
Ngubo
M+B
Maize (hand hoe)
yes
3
Ofafa
Velephi
Hadebe
M+B
Maize (hand hoe)
yes
3
Madzikane
Vakashile
Gambu
M+B
Maize (hand hoe)
yes
3
Cosmos
Xaba
M, M+B
Maize (hand hoe)
yes
4
Ngongonini
Mandla Mkhize
Mkhize
M+B
Maize (hand hoe)
yes
3
Mayizekanye
Babhekile
Nene
M+B
Maize (hand hoe)
YES
3
Manene
Mkhize
M+B
Maize (hand hoe)
YES
3
Gobizembe
Rita
Ngobese
M+B + SCC
Maize (hand hoe)
YES
4
Ozwathini
Doris
Chamane
M+B
Maize (hand hoe)
yes
3
Martina
Xulu
M+B
Maize (hand hoe)
yes
3
Aron
Nkomo
M+B
Maize (hand hoe)
YES
3
Soil health scores
Three assumptions are made regarding SH scores:
SH scores for the CAtrial plots will be higher than for the conventionally tilled control
plots.
SH scores will increase over time for CA trial plots.
SH scores for different cropping combinations, such as mono cropped plots, intercropped
plots and multi cropped plots will be different.
Soil health assessments over time
BERGVILLE
As this area is themost well-established with on-farm trials, soil health assessments have been
conducted for a number participants over four seasons. To determine trends in soil health over
time, theresultsfrom all CA trial plots (M, M+B, M+C,Dolichos, SCC) foreach soil health parameter
were averaged for 3 participants across the Bergville study area.
The figures for theveld benchmark samples have not been included. The results for these samples
mirrored the same trends as the CA trial plots although to a lesser extreme.
The participants are:
Dlezakhe Hlongwane (Stulwane):2015-2020
Ntombakhe Zikode (Eqeleni):2018-2020
Phumelele Hlongwane (Ezibomvini): 2016-2020
The figure below shows the result of this analysis.
Figure 7: SH scores between 2015/16 and 2020/21 for 3 participants in Bergville
From the above figure the following trends are visible:
The soil health index scores vary considerably between 2015-2020. The first reduction is
from 2015-2016 the latter following an extreme short term drought year. After that for
2017-2018 the SH scores increased substantially (16,6 and 17,9 respectively) to a similar
level asthe starting point in 2015 (16,5), just to plummet again for 2019and 2020. The
latter two years showed very late onset of rains, with increased heat and midseason dry
spells.
CO2 - C organic NOrganic CC/N%SOM
Soil Health
Calculation
(Index)
Bergville 2015179.1 7.489.0 12.116.5
Bergville 201675.113.6 205.1 15.73.18.4
Bergville 2017100.8 21.1 302.3 13.43.316.6
Bergville 2018147.6 15.7 233.2 16.73.517.9
Bergville 201965.820.8 260.9 12.83.513.9
Bergville 202055.56.6146.5 28.53.19.1
0.0
50.0
100.0
150.0
200.0
250.0
300.0
350.0
SH parameters for 3 farmers from Bergville 2015-2020
Organic N increased at variable levels from 2015 to 2019and then decreased dramatically
for the 2020 season.
Organic C increased although at variable levels between 2015-2019and then also reduced
substantially for 2020.
CO2-C respiration has shown the same periodicity as theoverall SH scores, decreasing
dramatically after a drought, late onset rains and mid-season dry spells.
% Soil organic matter (SOM)has remained reasonably stable as expected buthas shown
no overall increase in the 6-year period of CA implementation.
If one now looks at the SH scores over time for a selection of individual farmers in Bergville, from
three different villages (Ezibomvini, Stulwane and Eqeleni), one can see the sametrends as
indicated above mirrored for each individual, despite the individual differences and variations.
Figure 8: Comparison of SH results over time for 3
individuals in Bergville.
The extreme climatic conditions in the area,
including heat and dry soil profiles, reduces the
soil health impact of the CA practices and
increases variability in theresults for different
seasons. The analysis in effect becomes an
analysis of the effect of climate changeon the
soil health indicators in the area.
2015 2016 2017 2018 2019 2020
organic N7.415.3 19.0 14.9 20.75.4
Organic C89.0214.0 309.0 257.5 286.8 149.4
C/N 12.1 14.0 16.3 17.6 13.8 29.6
Soil Health
Calculation (Index)16.5 9.6 17.422.614.9 9.1
CO2 - C179.182.3111.1 211.071.055.8
%SOM 3.6 3.1 3.4 3.9 3.0 3.2
0.0
50.0
100.0
150.0
200.0
250.0
300.0
350.0
SH results forDlezakhe Hlongwane
(Stulwane) from 2015 to 2020
2016 2017 2018 2019 2020
organic N11.8 23.6 16.0 22.06.3
Organic C196.2 296.6 225.0 237.9 154.8
C/N 17.4 11.5 14.1 11.1 29.3
Soil Health
Calculation (Index)7.213.2 12.6 12.69.1
CO2 - C67.8 54.9 69.2 56.4 53.2
%SOM 3.1 3.2 3.4 4.3 3.1
0.0
50.0
100.0
150.0
200.0
250.0
300.0
350.0
SH results forPhumelele Hlongwane
(Ezibomvini)from2016-2020
2017 2018 2019 2020
organic N20.7 16.1 19.78.1
Organic C301.3 217.0258.0 135.4
C/N 12.4 18.4 13.4 26.7
Soil Health
Calculation (Index)19.1 18.5 14.39.2
CO2 - C136.4 162.670.057.7
%SOM 2.8 3.2 3.3 3.0
0.0
50.0
100.0
150.0
200.0
250.0
300.0
350.0
SH results forNtombakhe Zikode
(Eqeleni) for 2017-2020
Looking more closelyat one of the participants, Phumelele Hlongwane from Ezibomvini and
comparing the SH results for different plots within her 10block CA trial, one can see that even
within one field there can be quite a lot of variability in the SH results.
The table below indicates crops and crop combinations planted in each plot between 2017 and
2020 and gives an indication of the rotations Phumelele has used.
Table 10: Crop rotations in the 10 plots of Phumelele Hlongwane’s 10 block (1000m2) Ca trial plot over time (2015-2020)
SH samples were taken from the plots marked in orange in the above table between 2016 and
2020. The results for 2017-2020 for each plot number is shown below.
Table 11: SH parameter results for Phumelele Hlongwane 2017-2020, according to CA trial plot numbers
CA trial plot numbers
SH parameters/ per year
(2017-20200
2
4
5
6
9
10
CA
Contrl
Veld
Overall
average
Average of Soil Organic Matter
2017
3,2
3,2
3,2
3,4
4,2
3,3
3,4
2018
3,5
3,6
3,6
3,2
3,5
2019
3,7
3,6
3,9
3,5
4,0
2,9
4,1
3,7
2020
3,1
3,2
3,1
3,4
2,9
3,0
3,1
Average of CO2 - C
2017
42,6
68,7
62,3
68,8
61,2
97,3
66,8
2018
82,4
48,0
25,0
68,5
56,0
2019
61,1
48,0
40,5
52,8
62,0
53,7
76,4
56,4
2020
53,2
27,2
50,2
62,1
76,3
50,1
53,2
Average of Organic N
2017
19,5
6,5
17,8
19,9
17,8
16,1
16,3
2018
17,0
16,0
14,0
14,7
15,4
2019
28,6
23,6
16,8
23,0
28,9
16,0
17,1
22,0
2020
2,5
8,9
6,9
5,7
5,8
8,1
6,3
Average of Organic C
2017
232,0
157,0
235,5
256,5
227,5
260,5
228,2
2018
197,0
211,0
187,0
237,0
208,0
2019
280,0
219,0
191,0
199,0
328,0
227,0
221,0
237,9
2020
147,0
113,0
198,0
163,0
185,0
123,0
154,8
Average of C/N
Plot no
2015/16
2016/17
2017/18
2018/19
2019/20
2020/21
Run off plots
1
M+B
M
M +WCC
SCC
M
M+B
Orange= SH sampling
2
SCC
M
M+B
M+CP
M+B
SCC
Rotations have been
done attempting to
ensure a different
crop/crop mix on
each plot in each
consecutive year.
3
M+SCC+WCC
M+B
M
M+CP
M+B
M
4
M+B
LL
M
M+B
M+CP
M+B
5
LL
M
LL
M
LL
M
6
M+LL
SCC
M+CP
M+B
SCC
M
7
M+CP
M
M+CP
M+B
M+B
M+CP
8
M+B
M+CP
B
M+B
M+B
M+PK
9
M+CP
M+B
SCC
M
M+B
LL
10
M+B
M+B
M
LL
M
M+B
Control: M
(CA)
CA
Control: M
CA
Control: M
CA Control M
CA Control
M
CA control
M
CA
Control:
M+B (CA)
Conventional
control: SP
CV control
M
2017
12,7
24,2
13,2
14,5
13,5
15,9
15,7
2018
12,0
13,0
14,0
16,1
13,8
2019
9,8
9,3
11,4
8,7
11,3
14,2
12,9
11,1
2020
58,8
12,7
28,7
28,6
31,9
15,2
29,3
Average of Soil Health
Calculation (Index)
2017
10,9
10,7
12,7
14,0
12,5
16,5
12,9
2018
13,8
11,0
7,6
13,1
11,4
2019
14,6
11,5
9,6
11,6
15,7
11,5
13,8
12,6
2020
8,5
5,9
9,7
10,0
11,9
8,3
9,1
Originally the intention was to see whether different crops and crop combinationspositively
affect soil health in any given season (between different plots) and over aperiodof time on the
same sampling point or plot. It was however foundthat theinherent abioticsoil characteristics,
especially soil texture, has a much larger impact on the soil health results than the crop
combinations in a given season when different plots are compared. In other words, the variation
in soil characteristics between plots has a much bigger effect on soil health than different
treatments.
We hypothesised that over time given the multi-cropping rotations, that the soil health test
results on each plot would improve and also that the SH test results for the different plots would
slowly become more even, or more closely resemble each other thus lessvariation between
plots.
Figure 9 shows that there is a large variation in the active soil health parameters over time and
space (between thedifferent plots), i.e. average of CO2 (respiration), Organic N and C, C:N ratio
and even theSoil HealthIndex. The only stable soil health parameter, i.e., SOM, shows steady
increases per plot for three years (2017-2019), however, all plots had a fairly sharp drop in 2020,
which is unexplainable. Even the Veld plot had a sharp drop in SOM from 4,1 to 3, which can’t be
explained, unless there was either adifferent spot sampled with higher clay%, or an errorin the
laboratory procedure. Similar sharp increases in 2020 are also seen with organic C:N ratios in all
the plots.
Basically, the variability in plot level SH parameter results, especially the active parameters, have
shown the same trends as the SH resultsaveraged across the whole CA trial plot and across
participants in and between villages in Bergville. This then provides more evidence for the
argument that thevariability is a result of climate variability, rather than any specific farming
practice or intervention.
From the above figure the overall trends in increase anddecrease for thedifferent SH parameters,
mirrors that of Figure 7 above. The rather dramatic decrease in Organic N and to a lesser extent,
Organic C in the 2020 season isevident. Overall SOM values show a steady increase over three
years (2017-2019) and then aslight drop in 2020; the reason behind this change in SOM is still
being investigated.
The high loss of Organic N and Organic C from the soil between 2019 and 2020 can be considered
a combined effect of hightemperatures early inthe season, leaching, due to extreme rainfall
events mid to late season (January-March 2021) and reduced stover or soil cover, due to
increased grazing pressure from livestock because of dwindling grazing in the area.
This effect, combined with the significantly lower CO2-C respiration values for 2020, indicate the
need fornew applications of stableorganic matter intothe system andNitrogen supplementation.
Despite gains made in soil health and structure through implementation of CA principles these
results clearly indicate that:
1.Smallholders need to ensure increased permanent soil cover through reduced removal of
stover during the winter season and
2.They need to increase and consolidate their multi-cropping processes- specifically with
legumes, the latter inter alia aiming to reduce the C:N ratio.
Without these interventions the effects of climate variability can easily outweigh the gains made
in their CA implementation.
MIDLANDS
To determine trends in soil health scores over time,the results from allCA trial plots for 4
participants were averaged across villages for the study area. The figures for the veld benchmark
samples have not been included. The results for the veld samples mirrored the same trends as the
CA trial plots although to a lesser extreme.
The figure below shows the result of this analysis.
Figure 9: SH scores between 2018/19 and 2020/21 for 4 participants in the Midlands region.
From the above figure the following trends are visible:
The soil health index scores have remained reasonablysimilar, and overall values are
fairly high (above 7 in most cases). These are also reflected inthe low microbial
respiration, and organic N values. Organic C values are moderately high.
2018 2019 2020 2018 2019 2020 2020
GobizembeMayizekanye Ozwathini
Average of Soil Organic Matter5.4 7.1 7.6 6.5 8.4 7.4 8.1
Average of C/N14.6 8.9 41.413.4 9.7 14.8 9.6
Average of organic N11.117.0 4.29.5 16.0 6.4 12.0
Average of Organic C163.0 152.0 126.0 126.5 150.0 119.0 150.0
Average of CO2 - C32.1 50.7 50.2 24.9 32.0 40.3 29.3
Average of Soil Health Calculation
(Index) 7.6 9.8 8.0 6.0 8.0 7.5 7.5
0.0
20.0
40.0
60.0
80.0
100.0
120.0
140.0
160.0
180.0
SH scores for 4 participants from the Midlands from 2018-2020
A similar dramatic reduction in organic N, as was seen for Bergville above, also occurred
in Gobizembe forthe 2020samples. This isrelated to the sharp increase inC:N ratio in
the same region.
SOM values are generally very high in the Midlands area. There is afairlysharp increase
in % SOMunder CA in Gobizembe (2018-2020) and Mayizekanye, but a slight drop in
2020 in the latter region, although still very high at 7,4%.
The low microbial respiration is indicative of soils withvery little microbial activity. The soils in
this area are naturally high in organic matter, more so than for Bergville for example and these
low respiration values are indicative of long termbiological mining’ or suppression of the soils
frequent disturbance through tillage with too little diversity through cropping systems and living
roots and good soil organic cover. The fairly high Organic Cvalues is encouraging and shows a
fresh stream of food that could reactivate the soil microbes.
Recommendationsfor increased permanent soil cover and much increased inclusion of legumes
and cover crops are important here as well.
SOUTHERN KWAZULU NATAL(SKZN)
This area consists of reasonably spread-out villages from Ixopo to Creighton in SZKN.
To determine trends in soil health scores over time, the results from 5 participants on all CA trial
plots were averaged across villages into one average for each of the soil health parameters in the
study area. The values for the veld benchmark samples have not been included. The results for
the veld samples mirrored the same trends as the CA trial plots although to a lesser extreme.
The figure below shows the result of this analysis.
Figure 10: SH scores between 2017/18 and 2020/21 for 5 participants in SKZN
From the above figure the following trends are visible:
2017 2018 2019 2020
Average of CO2 - C, ppm C92.0 119.8 78.253.4
Average of Organic C ppm C175.5 172.9 172.5 129.7
Average of Organic N ppm N15.0 12.5 13.77.5
Average of C:N ratio11.7 13.9 13.9 53.4
Average of Soil health Calculation13.2 14.7 12.68.7
Average of %OM5.7 9.2 6.3 6.3
0.0
20.0
40.0
60.0
80.0
100.0
120.0
140.0
160.0
180.0
200.0
SH scores for 5 participants from SKZN 2017-2021
The soil health index scores here follow a similar trend as those in Bergville increasing
for 2017-2018, decreasing in 2019 and with a dramatic decrease for 2020
This dramatic decrease for 2020 is related primarily to asignificant drop in the Organic
N in the soil, as wellas Organic C, again atrend reflected in both Bergville and the
Midlands.
SOM% values show asteady increase under CA over time, although the high value in 2018
(9,2) is clearly an outlier due to possible errors in the laboratoryor during thesoil
sampling.
The drastic changes ofsensitive indicators, such as Organic N (related to a drop in Organic C and
CO2respiration) from the soil between2019 and 2020 can be considered a combined effect of
increasedtemperatures especially earlyin the season (September-November), extreme rainfall
events mid to late season (January-March 2021) and reduced stover or soil cover, due to
increased grazing pressure from livestock because of dwindling grazing in the area.
In SKZN however the microbial respiration figures are above average and high from 2017 to 2019
(92, 119, 72) when compared to the Midlands study area and are comparable to valuesin the
Bergvillestudy area. The 2020 result of 53,4ppm, which is slightly above average,however,
indicates the need for food for the soil food web (SFW), which is best derived from healthy diverse
cropping systems and living roots in the soil, as well as soil organic cover. The practices or factors
were probably negatively influenced by environmental factors, such as drought and high
temperatures.
Overall, the results for all three areas indicate that climate variability (short term droughts
late onset of summer rains, mid-season dry spells and high early season temperatures)
had a significant effect onsoil health status and that unless CA participants can
substantially improvetheir soilcover and crop diversification with leguminous cover
crops, gains made by the present level of reduction in soil tillage, increase in diversity and
small increases in soil cover,are notlikely to beable to sustain the higher yields that
participants are looking for.
Different CA cropping options
The CA experimentation process has been designed to maximise crop diversity.
The following progression has been used:
Years 1 and 2: Single cropped plots of maize (M) and beans (B) and intercropped plots of
M+B and M+C (cowpea).
Year 3: Inclusion of cover crops; a 3-species mix of summer cover crops (SCC) including
sunflower, millet andSun hemp and a 3-species mixof winter cover crops (WCC),
Saia/black oats, fodder rye and fodder radish.
Year 3+: Inclusion of legume cover crop Lab-lab beans (LL).
Year 3+: Rotation of the above-mentioned plots within the CA trial (ten plots).
Year 3+: Inclusion of permanent fodder strips and strip-cropping as a crop rotation
strategy.
The assumption is that the combinationof multi-cropping and crop rotation would provide for
the fastest build-up of organic matter and improvement of SH for this smallholder CA system. The
assumption was also made that if crop rotation is included as a practice in such a multi-cropping
system, then the SH for all the plots wouldincrease over time and that variability between the
plots would decrease, as each plot undergoes a rotation of multiple crops.
SH samples were taken for 13 Participants in their 5th to 8th year of CA implementation who have
used both multiplecropping and crop rotation in their CA plots to ascertain whether these
assumptions could be proven. Samples have been taken from the sample plots every season, even
though the crop on each plot differs in each season. The figure below is indicative.
Figure 11: SH scores for different cropping options for 13 participants in Bergville for 2020/21
From the above figure the following trends can be seen:
The SH scores, althoughlow for this season as explained in the previous section are all similar
and additionally are also similar to the SH index scores ofthe veld samples. Ifthese average values
are used, theassumption that SH scores in multi-crop rotations will become similar over time can
be verified as can the assumption that SH scores will improve and more closely resemble the veld
benchmark scores.
However, variability for SH scores between different plots in individuals’ CA trials is high due to
variation in inherent soil characteristics, especially texture, climate, and the implementation of
CA practices bythe participants. From the approach taken here, the major factor in the change
(reduction) of soil health scores seems to be climate variability. Furthermore, it isatthis stage
not possible to separate the effect of cropping systems on soil health, compared to climate
variability and the inherent soil characteristics. Another more comprehensive methodological
Bean DolichosM+BMaizeMaize
control SCC Veld
Average of organic N5.4 5.7 9.4 7.6 8.5 6.6 9.6
Average of Organic C131.0 163.0 144.3 138.0 156.1 133.8 163.8
Average of C/N24.3 28.6 19.6 22.1 31.2 24.8 29.2
Average of Soil Organic Matter5.2 3.4 4.2 4.5 4.1 4.1 4.5
Average of CO2 - C42.1 62.1 45.7 44.7 48.4 50.3 57.3
Average of Soil Health Calculation
(Index) 7.4 10.0 8.48.18.98.4 10.0
0.0
20.0
40.0
60.0
80.0
100.0
120.0
140.0
160.0
180.0
SH scores for differentcropping options within a CA rotation for 13
participants in Bergville (2020/21)
approach, such as a multivariate analysis of all the factors and parameters per participant, might
be necessary to improve the analyses.
Soil organic matter (SOM)
Soil organic matter is where soil carbon is stored and is directly derived from biomass of
microbial communities in the soil (bacterial, fungal, and protozoan), as well as from plant roots
and detritus and biomass-containing amendments like manure, green manures,mulches,
composts and crop residues (Motaung, 2020). Soil organiccarbon refers to the carboncomponent
of soil organic matter (SOM) that is measured in mineral soil which passes through a 2 mm sieve.
It is the largest component of SOM (approximately 45%) and the easiestand cheapest to measure.
The SOC content of agricultural soils is generally between 0,5% and 4%.
There isgeneral agreement among researchers that CA improves soil health by increasing soil
organic matter (SOM). Increased soil organic carbon (SOC) also sequestrates atmospheric carbon
and thereby contributes to climate change mitigation targets (Swanepoel, et al., 2017).
Despite this understanding, some studies in Africa and Southern Africa have reported only small
increases, or very slow build-up of SOC and even negligible increases in some cases, compared to
conventional systems. It is being argued that the build-up of SOM in CA systems is relatedmore
to increased biomass and organic mulch than a reduction in tillage (Giller, et al., 2009). It has also
beensuggested that in drier climates with sandier soils it may takeup to ten years to detect a
noticeable difference in SOC accumulation (GRDC, 2014).
In addition, it has been suggested that to get reliable data on changes in SOC related to land use
patterns, largenumbers of samples need to be takenbetween 80 and 500 on average for
croplands and undisturbed soils such as forests, respectively (Stolbovoy, et al., 2007). This is not
financially feasible for small, multifaceted programmes such as the CA-SFIP.
Methods
Sampling for laboratory analysis of soil health is, as described above in section 1, with 20
subsamples making up one composite sample within a100 m2plot, taken during the fallow
season (August-September each year) andair dried and stored at room temperature prior to
analysis. Soil sampling depthwas 0-10cm, as this is the depth ofsoilwhere the most changein
SOC is likely to occur.
PercentageSOC (kg C/kg soil x 100) has been provided for these samples (Soil Health Solutions
Laboratory, WC). A constant factor of 1,72is usedin South Africa, to convert%SOC to %SOM
(GRDC, 2014).
BERGVILLE
Percentage SOM has been calculated for participant samples between 2018-2020 to ascertain
whether there is a build-upof carbon as an outcome of theCA implementation for participants.
The average results for all theCAmulti-cropped plots (both intercropping and crop rotation)of
the 7-13 participantsfor whom samples were taken in this timeperiod, is presentedin the table
below, along with the average of theControl plots (both CA and conventional controls mono-
cropped over time with maize) and an average of the veld samples.
Table 12: %SOM and tons of Carbon/ha for Bergville CA and Control plots: 2018/19 to 2020/21
%SOM
tC/ha
2020
2019
2018
2020
2019
2018
CA -multi-cropping
4,05
4,75
4,13
3,1
3,6
3,1
Control M*
3,79
4,79
4,46
2,9
3,6
3,4
Veld
4,75
4,62
5,66
3,6
3,5
4,3
*This value is a combination of control plots, which included CA control plots as well as conventionally tilled plots planted to maize only
in consecutive years. This combination of controls was done for this analysis only, to get an averaged OM value for maize only plots as
a control compared to multi-cropped CA plots
As with the soil health samples for 2020, there has been a reduction in SOM measured in the soil
for 2020 for the CA plots and thecontrol plots, when compared to the2019results. This is very
unusual, since SOMdoesn’t not easily change over afew years, especially if the land use or
management stays constant. % SOM in the veld samples increased slightly for 2020, when
compared to 2019.
This result provides further urgency to the need for improving soil cover and crop diversification
by the smallholder participants, to offset the effects of climate variability in the system.
MIDLANDS
Percentage SOM has beencalculated for participant samples between 2018-2020 to ascertain
whether there is a build-upof carbon as an outcome of theCA implementation for participants.
The average results for all the CA plots of the 4 participants for whom samples were taken in this
time period, is presented in the small table below, along with the average of the CA Control plots
and an average of the veld samples.
Table 13: %SOM and tons of Carbon/ha for Midlands CA and Control plots: 2018/19 to 2020/21
%SOM
tC/ha
2020
2019
2018
2020
2019
2018
CA multi-cropping
7,8
7,5
5,1
5,9
5,7
3,9
CA Control M
8,4
8,3
6
6,3
6,3
4,5
Veld
8,2
10,2
8,1
6,2
7,7
6,1
For both the CA trialplots (M,M+B, M+CP and SCC) and the CA control plots (maize monoculture)
the %SOM has increased in both 2019 and 2020 when compared tothe 2018 result. This indicates
the build-up of SOM through the CA processin theMidlands, at a rate of 0,2t/ha/ annum on
average higher for the CA plots (both trial and control) when compared to the veld samples over
the time period tested.
SKZN
Percentage SOM has beencalculated for participant samples between 2018-2020 to ascertain
whether there is a build-upof carbon as an outcome of theCA implementation for participants.
The average results for all the CA plots of the 5 participants for whom samples were taken in this
time period, is presented in the small table below, along with the average of the Control (both CA
and conventionally tilled plots planted to mono cropped maize) plots and an average of the veld
samples.
Table 14: %SOM and tons of Carbon/ha for SKZN CA and Control plots: 2018/19 to 2020/21
%SOM
tC/ha
2020
2019
2018
2020
2019
2018
CA multi-cropping
6,7
6
5,7
5,1
4,5
4,2
CA Control M
7,1
5,7
5,9
5,4
4,2
4,5
Veld
7,4
8,3
8,0
5,6
6,3
6,0
Note: the sharp increase in %SOM for the control M plots in 202 is physically almost impossible to obtain in one season for
one parcel of land. This value is an average of the %SOM for 5 participants an although extremely unusual needs tobe
considered asa valid result. As discussed in the section above it is suspected that the high levels of change of Organic N
and Organic C in the soil, due in the most part to extreme weather conditions is at the root of these unusual outcomes.
For theboth theCA trial plots (M+B and M+C)and the control maize plots the %SOM has
increased quite sharply for both 2019 and 2020 when compared to 2018. The %SOM for the veld
samples have however been variable, first increasing and then decreasing in the two consecutive
seasons since 2018. This indicates the build-upof OMthrough the CA process in SKZN, at arate
of 0,6t/ha/ annum on average higher for the CA plots (both trial and control) when compared to
the veld samples over the period tested.
Overall, the %OM results indicate that the practice of CA is important in providing for
build-up of organic matter in the soil, at average rates of -0,1t/ha/annum (Bergville,
0,45t/ha/annum (SKZN)and 0,7t/ha/annum (Midlands). The impacts of climate
variability, specifically in Bergville, but also in the other areas has negated much of the
potential positive impact of organic matter build-up necessitating a far stronger emphasis
on crop residues and crop diversification than ispresently the case to mitigate for this
effect.
Mycotoxin analysis
Mahlathini is collaboratingwith another project funded by The Maize Trust and managed by Dr
Belinda Janse van Rensburg from the ARC Grain Crops Institute in Potchefstroom to analyse the
presence of mycotoxins on smallholder CA participants’ maize.
Mycotoxins are secondary metabolites produced by filamentous fungi which contaminate a large
fraction of the world’s food, mainly staple foods such asmaizeand beans in our case. This
worldwide contamination of foods is an enormous problem to human populations, principally in
less industrialized countries and in the rural areas of some developed countries. The adverse
effects of mycotoxins on humanhealth can beboth acute and chronic, provoking problems such
as liver cancer, reduction of immunity, alterations in the protein metabolism, gangrene,
convulsions, and respiratoryproblems, among others. Mould growth canoccur either before
harvest or after harvest, during storage and on/in the food itself often under warm, damp and
humid conditions. Most mycotoxins are chemically stable and survive food processing.
Several hundred different mycotoxins have been identified, but the most commonly observed
mycotoxins that present a concern to human health and livestock include aflatoxins, ochratoxin
A, patulin, fumonisins, zearalenone and nivalenol/deoxynivalenol.
Some factors which influence the presence of mycotoxins in foods are related to environmental
conditions, such as storage, that can be controlled without too much expense.
Goodpractices for reducing the presence of mycotoxins aresummarized inthe following short
list.
Host resistant or tolerant varieties are the most cost effective and practical means of
combating the disease.
Avoid planting maize at unacceptably high densities as this increases stress.
Rotate with non-hostsof the Fusarium graminearum species complexsuch as legumes,
cotton and sunflower.
Harvest early to avoid lodging. Get rid of infected debris to avoid build-up of inoculum.
Control insects such as stalk borerswhich may serve as possible vectors, observingthe
threshold value of 10% infested plants for chemical control.
In order to prevent ear rot after harvest, store grain under low moisture conditions.
Maize that is harvested can befurther decayed and contaminated by thealready
present Fusarium spp. Maize harvested must contain a low moisture content and the area
in which you store the maize must be kept dry and clean.
The table below summarizes the mycotoxins tested for and some of their characteristics.
Table 15: Mycotoxins tested for and health effects in humans and animals.
Mycotoxin type
Fungal origin
Health effects in Humans
Health effects in animals
Aflatoxins (B1,B2
G1, G2)
Aspergillus
flavus and
parasiticus
Liver damage, live cancer,
immunosuppressive,
damage to DNA - genotoxic
(10 µg/kg)
Liver cancer, immunosuppressive
(50 µg/kg)
Fumonisins
Fusarium spp,
Fusarium
verticillioides
Oesophageal cancer in
humans, liver toxicity,
immunosuppressive (2000
µg/kg)
Carcinogenic, liver and kidney
toxicity, immunosuppressive
(5 000 µg/kg for horses and pets,
10 000 µg/kg for pigs and 50 000
µg/kg for cattle and poultry)
Trichothecenes
(T-2 HT-2)
Fusarium spp
Liver damage, live cancer,
immunosuppressive
Liver damage, live cancer,
immunosuppressive
Deoxynivalenol
(DON)
Fusarium spp
Anaemia, skin lesions,
vomiting, diarrhoea, and
damage to liver damage
(2000 µg/kg)
Anaemia, skin lesions, vomiting,
diarrhoea, and damage to liver
damage
(1000 µg/kg for pigs, pets and
calves, up to 5000 µg/kg cattle)
Zearalenone
(ZEN)
Fusarium ssp
Hormonal imbalance and
reproductive effects
Carcinogenic, hormonal imbalance
and reproductive effects (500
µg/kg for cattle, 3000 µg/kg for
pigs)
Ochratoxin A
Aspergillus
and
Penicillium spp
Upper urinary tract disease,
carcinogenic, liver and cell
toxicity
Upper urinary tract disease,
carcinogenic, liver and cell toxicity
(50 µg/kg pigs, 200 µg/kg poultry)
NOTE 1: References: National mycotoxin regulations for foodstuffs to be consumed by humans (GovernmentGazette, No
987) and National mycotoxin regulations for maximum allowable levels ofmycotoxins in animalfeeds (Act 36of1947;
Government notices No.R.70; SAGL, 2019,)
NOTE 2: Neither aflatoxins nor Ochratoxin were found in samples analysed.
Fusarium ear rot isespecially common in fields with bird or insect damage to the ears. Affected
ears usually have individual diseased kernels scattered over the ear or in small clusters
(associated with insect damage) among healthy-looking kernels. The fungus appears as a whitish
mouldand infected kernels sometimes develop a brownish discoloration with light-coloured
streaks (called starburst)
Gibberella ear rotis similar but presents more as a white- pinkish mould. The main difference is
that it is favoured bycool, wet conditions whereas the fusarium ear rot isfavoured by hotand dry
conditions. Gibberella and fusarium ear rot pathogens overwinter on corn residue and in the soil.
(From: www.pioneer.com/us/agronomy/common_corn_ear_rots_cropfocus.html)
Below is asmall table outlining the participants and samples where high levels of mycotoxins
were found.
Table 16: Mycotoxin presence for samples taken in Bergville, Midlands and SKZN, 2019/20
Area
Village
Name and
Surname
Plot
Descriptio
n
% Ear
rot
Fumonisin
µg/kg
DO
N
µg/
kg
15-
Acet
yl-
DON
µg/k
g
ZEN
µg/
kg
T2
µg/
kg
HT-
2
µg/
kg
B
1
B
2
B
3
Midla
nds
Mayizek
anye
Fikile
Maphumolo
Conv Maize
30
1826
127
313
Fikile
Maphumolo
CA Maize
3
976
Nomusa
Shandu
CA Maize
12
4719
1233
944
Nomusa
Shandu
M+Cp
14
587
1409
1446
Babhekile
Nene
M+Cp
25
539
174
88
4294
SKZN
Madzika
ne
Cosmos Xaba
M+B, M+Cp
0
2050
Bergvi
lle
Ezibomvi
ni
Phumelele
Hlongwane
CA Maize
1
157
77
1950
Phumelele
Hlongwane
M+SCC
0
183
57
1418
Stulwane
Neliswe Msele
CA Maize
0
466
92
57
1665
Red cells indicate toxic levels of themycotoxin concerned and pink cells indicate high levels,
although slightly below the maximum level allowable.
A total of 36 samples were taken from 10 participantsacross all threeareas. Of these 9 samples
contained high levels of mycotoxin and 10 samples contained no mycotoxins.15 Samples
contained low levels of a range of mycotoxins. All mycotoxins wereproduced by Fusarium spp.See
appendix 1 for the full analysis. In a number of cases participants had both high levels of
mycotoxins in some of their plots and no mycotoxins in others.
In general mycotoxin levels inmaize for CA plots intercroppedwith beans werelow and
much lower than some of the maize only plots. The results indicate an urgent need to work
with smallholder farmerson strategies to reduce mycotoxin levels intheir maize, both in
their fields and in subsequent storage methods and processes.
Progress summary
Monitoring results were summarized in the six-monthly report. In this section we provide a
summary of monitoring for trails planted in late January 2021 and for the experimentation in
cover crops, and livestock and poultry fodder production
Collaboratively managed trials (CMTs).
BERGVILLE
Compiled by Lungelo Buthelezi, Nkanyiso Mzobe and Michael Malinga
For this adaptive research process, we have focused mainly onfourvillages: Stulwane,
Ezibomvini, Vimbukhhalo and Eqeleni with a total of 64 participants.
Strip cropping (2m wide strips w 4 lines of crop per strip) on either 400m2or1000m2plots was
introduced as an alternative to the 100m2 blocks used to date. Participants have found this layout
a lot easier to handle and implement and many of the participants thus opted for strips.
Bergville additional fodder, poultry feed and seed plots
Ten (10) participants across 3 villages(Stulwane, Ezibomvini and Eqeleni)undertook the
experimentation in fodder and cover crop seed plots, over and above planting their 1000m2 block
and strip trials. Three participants in Vimbukhalo who undertook to do this experimentation
(ZweniNdaba, Sphelele Zondo, NomusaZikode)found theadditional load of work too onerous.
In addition, 5 participants in Stulwane volunteered, but never planted (Nikeziwe Ndlovu, Thembi
Nsele, Mpithi Mabaso, Mrobeth Miya and Khethabahle Miya).
Of the participants who planted thefodder strip cropping plots, the trials for Sabelo Mbhele and
Dombolo Dlamini (Stulwane), Ntombenhle Hlongwane (Ezibomvini)and Thulile Zikode (Eqeleni)
were destroyed due to livestock invasions.
For the seed production plots, participants were supplied with fencing material to fence a 250m2
area within which to produce the cover crop seed. The idea was to plant the cover crops and
Dolichos in theseareas. Four participants received the fencing: Phumelele Hlongwane, Landiwe
Dlamini and NtombenhleHlongwane (Ezibomvini) and Zweni Nadaba (Vimbukhalo).
Ntombenhle has beenill with COVID-19 and did not plant. Zweni Ndaba’s fencingwas returned
to MDF as she was also ill and did not plant.
This lack of implementation has been as a direct result of COVID-19, where 3-4 of the participants
were very ill and others were unfocused due to increased pressure on their household incomes
and livelihoods. Livestock which are usually herded or moved to the mountains in summer, have
remained in the villages and have caused substantial crop damage. Below a summary is provided
for some of the participants.
Stulwane
Thulani Dlamini
He planted the following trials:
Year
Date of
planting
Size
of
trial
Trial
type
Plot 1
2
3
4
5
6
7
8
9
10
2018/2019
17/11/2018
1000
blocks
SCC
M+B
M+C
B
M+C
M
CP
M
B
M
2019/2020
18/12/2019
1000
blocks
LL
SCC
M+B
B
B
M
B
SCC
M+C
LL
2020/2021
17/11/2020
1000
blocks
M+PK
CP
M+C
LL
SCC
B
M+B
M
M+B)
M+B
16/12/2020
1000
CA
Strip
M
B
M
CP
M
SCC
M + SCC
M
C
M
26/02/2021
371
Poultr
y feed-
strips
SCC
M
SCC
M
Note: Maize varieties; (blocks) PAN53 and (strips) PAN5A190. Beanvarieties:PAN9292.
Pumpkin varieties: Queensland Blue. SCC varieties: Sun hemp, sunflower, fodder sorghum.
Figure 12: Above left and right: The maize and SCC trips planted in late February 2021 as poultry feed.
Comments:Mr Dlamini planted both the SCC and PAN5A190short season yellow maize, to be
able to make amixed feed for hissmall broiler business. The trial was growing well, but there was
some danger of livestock invasions as he planted late in the season. He has however undertaken
to look after this plot and ensure harvesting of the seed.
Khulekani Dladla
He planted the following trials:
Year
Date of
planting
Size of
trial
Trial type
Plot
1
2
3
4
5
6
7
8
9
10
2018/2019
20/11/2018
1000
block
SCC
M+B
M+C
M+B
B
M
LL
M
B
M
2019/2020
14/12/2018
1000
block
M
SCC
M+B
M+C
M+
B
B
M
LL
M
B
2020/2021
18/11/2020
1000
strip
SCC
LL
M)
B
M+
C
C
M+
B
M+Pk
M+C
M+C
17/12/2020
1000
Strip
M
B
M
CP
M
SC
C
M
LL
M
B
End Feb
2021
182
Poultry
feed
strip
C
SCC
C
SCC
C
SC
C
C
SCC
Dec 2020
Fodder
strip
M
Luc
M
Pens
M
Lsp
M
Pens
M
Lsp
Note: Maize varieties; (blocks) PAN53 and (strips) PAN5A190. Bean varieties: PAN146. Pumpkin
varieties: Queensland Blue. SCC varieties: Sun hemp, sunflower, fodder sorghum. Fodder Lucerne
(Luc), Lespedeza (Lsp) and Pensacola (Pens).
Figure 13: Above left and centre: Lepedeza harvested and dried for baling as livestock fodder and Lespedeza seed collected
and Above Right: Khulekani Dladla standing in one of his late season SCC strip plots, planted as poultry feed.
Comments: For the fodder strip cropping, thelucerne took averylongtime to germinateand grow
and germination was not very good. Pensacoladidnot germinate at all. Thelespedeza germinated
and grew well and Khulekani has managed to harvest a reasonably substantial quantity of seed.
Nelisiwe Msele
She planted the following trials:
Year
Date of planting
Size of trial
Trial type
Plot 1
2
3
4
5
6
7
8
9
10
2018/2019
17/11/2018
1000
block
M+CP
M+B
M
CP
M
B
M
M
SCC
M
2019/2020
2019/12/12
1000
block
LL
M+CP
B
M
CP
M
B
B
M
LL
2020/2021
23/11/2020
1000
block
M
C
M+C
B
M
LL
M+B
M+PK
M+B
SCC
Jan 2021
100
Poultry feed block
SCC
Note: Maize varieties; (blocks) PAN6479. Bean varieties: Gadra. Pumpkin varieties: Queensland
Blue. SCC varieties: Sun hemp, sunflower, fodder sorghum
Figure 14: Right: Nelisiwe’s SCC plot
planted as poultry feed in January
2021. Fra right: Sunflower seed
harvested from this plot.
Comments: She has already
harvested the sunflower
seeds and is now waiting for
the Sun hemp and sorghum
to mature.
Nothile Zondi
She planted the following trials:
Year
Date of planting
Size of trial
Trial type
Plot 1
2
3
4
5
6
7
8
9
1
0
2018/201
9
2018/12/11
1000
block
SCC
M+
B
M+
C
B
M
LL
M
C
B
M
2019/202
0
19/12/2019
1000
block
M
C
SCC
M
B
M
LL
M
M+
C
B
2020/202
1
17/11/2020
1000
block
C
SCC
M
B
M+P
K
LL
M
M+
B
B
M
30/12/2020
1000
Strip
M
B
M
C
M
SC
C
M
LL
M
B
10/12/2020
300
Poultry feed block
SCC
January 2021
Fodder strip
M
Lsp
M
T
F
M
TF
M
Lsp
Note: Maize varieties; (blocks) PAN53and (strips) PAN5A190. Beanvarieties: PAN9292.
Pumpkin varieties: Queensland Blue. SCC varieties: Sunhemp, sunflower, fodder sorghum.
Fodder: Tall fescue (TF), lespedeza (Lsp)
Figure 15:Right:
The short season
maize and
perennial fodder
strip trial showing
tall fescue andFar
Right: Nothile’s
poultry feed block
of SCC.
Comment: Nothile has already harvested the sunflower from her poultry feed plot and is still
waiting for the Sun hemp and sorghum to mature.For her fodder strip trial, she has one strip of
Lespedeza that was planted in
2019/20 continuing this season
and has planted further strips of
Lespedeza and Tall Fescue. She
also planted Pensacola, but this
didnot germinate well. In future
the focus will be on lespedeza and
Tall Fescue grass.
Figure 16: Right and far right: Lespedeza
strips at different stages of maturity. She
kept one strip to go to seed, to keep for
future planting and usedthe other strip to
cut and store some fodder for her livestock.
The rest will be grazed directly by livestock
in the winter season.
Eqeleni
Ntombakhe Zikode
She planted the following trials:
Year
Date of planting
Size of trial
Trial type
Plot 1
2
3
4
5
6
7
8
9
10
2018/2019
06/12/2018
1000m2
block
M
SCC
M
B
M
LL
M+B
M+B
M+B
M+LL
2019/2020
05/12/2020
1000m2
Block
M + B
M
M + B
M
M + B
M
B
SCC
M + C
LAB LAB
2020/2021
17/11/2020
1000m2
Strip
M
B
M+Pk
SCC
M
C
M
B
M
Pk
100m2
Fodder block
Lsp+M
Note: Maizevarieties; (blocks) PAN6479. Bean varieties:
PAN9292. Pumpkin varieties: Queensland Blue.Fodder:
Lespedeza (Lsp) Pensacola (Pens), Maize PAN5a190
Comments: This season Ntombakhe did not maintain her fodder
plot well. Lespedeza however still managed to germinate and
grow despite high weed pressure.
Figure 17: Lespedeza growing among weeds in Ntombakhe’s fodder plot.
Ezibomvini
Zodwa Zikode
She planted the following trial plots:
Year
Date of planting
Size of trial
Trial type
Plot 1
2
3
4
5
6
7
8
9
10
2018/2019
21/11/2018
400m2
block
M+C
M+C
M
B
M+B
M
SCC
LL
M+B
M+B
2019/2020
2019/11/12
1000m2
block
LL
M+B
M+B
M+P
M+C
SCC
M
M
B
M
2020/2021
25/11/2020
1000m2
block
M
M
B)
M
M+C
M)
M+B
M
M
LL
Dec 2020
250m2
block
M
Teff
M
Lsp
Note: Maize varieties; (blocks) PAN53. Bean varieties: Gadra. Cover cops: Sunflower, Sun hemp,
fodder sorghum. Fodder: Lespedeza (Lsp) Teff, Maize PAN5a190
Comments: Zodwa harvested all the teff in her fodder plot and provided the fodder to her
neighbour Phumelele Hlongwane for her cattle.
Mantombi Mabizela
She planted the following trials:
Year
Date of planting
Size of trial
Trial type
Plot 1
2
3
4
5
6
7
8
9
10
2018/2019
21/11/2018
400m2
block
M
M+B
M
M+C
2020/2021
19/11/2020
1000m2
block
B
M
M
M+C
M
B
M+Pk
M
M
M+Pk
250m2
Seed block
Lsp
M
T
M
Note: Maize varieties; (blocks) PAN53. Bean varieties:Gadra. Pumpkin varieties: Queensland
Blue. Cover cops: Sunflower, Sun hemp, fodder sorghum. Fodder: Lespedeza (Lsp) turnip (T),
Maize PAN5a190
Comments: Mantombi’s
turnips did very well and was
cut in April 2021 to feed her
goats. Lespedeza has grown
very slowly and will be kept
for direct winter grazing as it
had not developed enough
biomass for cutting.
Figure 18: Right:Turnips in Mantombi’s
fodder plot which was harvested for
goats and Far Right: Lespedeza growing,
but suffering a bit from weed pressure.
Landiwe Dlamini
She planted the following trial plots:
Year
Date of
planting
Size of
trial
Trial
type
Plot 1
2
3
4
5
6
7
8
9
10
2018/2019
20/11/2018
400m2
block
M+B
M+C
M
LL
2019/2020
21/11/2019
500m2
block
M+B
M
M+C
M
B
2020/2021
24/11/2020
1000m2
block
M
M
M
M+Pk
M
LL
M
B
M
M+B
Dec 2020
420m2
fenced
block
sunflowe
r
sorgh
m
Sun
hemp
Sunflo
wer
Sorghu
m
Note: Maize varieties; (blocks) PAN53. Bean varieties: Gadra. Pumpkin varieties:Queensland
Blue. Cover cops: Sunflower, Sun hemp, fodder sorghum.
Comments: Shehas harvested her
sunflower seeds (0,428kg) and is still
waiting fortheSunhemp and
sorghum to mature.
Figure 19: Right and Far Right: Landiwe’s fenced
plot with summer cover crops planted for seed
production.
Phumelele Hlongwane
She planted the following trial plots:
Year
Date of
planting
Size of
trial
Trial type
Plot 1
2
3
4
5
6
7
8
9
10
2018/2019
2018/07/11
1000m2
block
SCC
M+C
M+C
M+B
M
LL
M
M+B
M+B
M+B
2019/2018
21/11/2019
1000m2
block
M
M+B
M+B
M+C
LL
SCC
M+B
M+B
M+B
M
2020/2021
16/11/2020
1000 m2
block
M+B
SCC
M
M+B
M
M
M+C
M+Pk
LL
M+B
16/11/2020
1000m2
Strip
M
B
M
SCC
M
Pk
100m2
block
M
PAN5A
190
250m2
Seed
block
M
PAN5A
190
T,
Luc
M
PAN
6479
SCC
Note: Maize varieties; blocks and strips PAN53. Beanvarieties: Gadra. Pumpkin varieties:
Queensland Blue. Cover cops (SCC): Sunflower, Sun hemp, fodder sorghum. Fodder turnips (T)
and lucerne (Luc).
Comments: For Phumelele’s 250m2 fenced plot for seed production she planted in 4 small blocks:
PAN 5A190 (short season yellow maize), fodder turnips and lucerne, PAN6479(left of hybrid
maize seed from last year)and SCCmix. She was not aware that one shouldplant the hybrid maize
for keeping seed. This seed will be used for poultry fodder as an alternative. The fodder species,
turnips and
lucerne did not
germinate or grow
well. As with other
participants, she
has already
harvested the
sunflower seed,
but is still waiting
for the Sun hemp
and sorghum to
mature.
Figure 20: Right: Phumelele’s strip cropping trial showing maize, cowpeas and SCC and far Right Her 250m2fenced plot for
cover crops seed production with Sun hemp visible in the picture.
Bergville bean yield summary and photos.
Compiled by Nkanyiso Mzobe
Bean yields were measured for 59 participants across7 villages (Ezibomvini, Eqeleni, Stulwane,
Ndunwana, Vimbukhalo, Emafefeteni and Thamela), for both the 10x10 CA intercropped plots
and the strip cropping plots. The two small graphs below provide a summary of the yield results.
Figure 21: Above Left: Comparison of block and strip planting bean yields, and Above right: Comparison of intercropped and
mono-cropped bean yields.
The average yield for beans in Bergville was 0,88t/ha, which is low. Yield reduction was primarily
due to heavy rains late in the season (March and April 2021), which led to a lot of water damage
in the drying beans anda reduction in yields. Participants were sure that the intercropping had
exacerbated this problem due to close spacing and shading from the maize intercrops. From the
small figure above right, it can however be seen that the average yield of intercropped beans was
1,59t/ha compared to 1,22t/ha for the single cropped blocks and 0,78t/ha for the single cropped
strip planting. Reasons for this difference are likely to be higher weed pressure inthe strip
plantings and also differences in inter-row spacing.
From the figure above left it can be seen also that theaverage yield for beans planted in the 10x10
blocks was higher than that of beans planted in the 4x25m strips.
In summary, bean intercropped with maize in the CA trials produce higher yields than the
mono-cropped beans, despite participants continuedbelief that this is not the case.
Participantsare also still reluctant to followthe closespacing regimes, insisting that it is
easier to weed when crops are more widely spaced, but not considering that this
concurrently increases the weed pressure.
Total
10X10 1.06
Strip 0.70
0.00
0.50
1.00
1.50
ton/ha
Bergville bean yields 2021.
Comparison between 10x10
block trials and strip cropped
trials
10X10 Strip
Average of
intercrop (t/ha)1.59
Average of
Monocrop (t/ha)1.22 0.78
0.00
0.50
1.00
1.50
2.00
Ton/ha
Bergville bean yields 2021.Comparison
between intercropped and
monocropped plots
Below are a few indicative photographs of bean weighing and yields for participants.
Figure 22: Left to Right: Phumelele Hlongwane (Ezibomvini), Bhukisile Mpulo (Vimbukhalo) and Lindiwe Bhengu (Thamela)
with their bean harvests for 2021.
MIDLANDS
Compiled by Temakholo Mathebula
For theMaiyzekanyevillages (x4) inSwaiymane, participantsgenerally plant alot later thanin
most of the other areas, opting to plant in mid-to end- January, rather than November,
necessitating a later monitoring date in March 2021.
Background on Mayizekanye Farmers
The community of Mayizekanye derives asignificant part of their income from various farming
activities, over and above small government social and pension grants, as well as to a much lesser
degree,remittances from young adults employed in the city. Their farming activities include but
are not limited to;maize, beans, sugar cane, amadumbe, sweet potatoes, avocadoes, poultry
(mainlytraditional chickens) and livestock farming. Most grow bothfor market and to
supplement household income andmajority of theparticipants working with MDF are between
the ages of 40 and 70 years old and are mostly unemployed women. Many households in the
villageare female headed mainly dueto husbands passing away, migrating to the city due to work,
men spending the majority of their time in local taverns and other undisclosed reasons.
Overview of the 2020/21 growing season.
This season the team introduced larger CA trials, in the form of 1000 m2 10x10 plots and 1000 m2
CA strips as well, which is more than double the standard400 m2CA trial size introduced inthe
area in thepast. Thirty-five (35) Participants were involved, with a substantial group ofnew
participants. The focus however has been on the CMT’s for the longer term participants.
Crop growth monitoring was carried out during the month of March 2021and all participants
were monitored. The nine longer term participants (Mrs Nxusa, Babhekile Nene, Nakeni Ngubane,
Fikile Maphumulo, Fikelephi Maphumulo, Florence Cebekhulu, Thembi Mkhize, Lungile Madlala,
Ntombi Shandu and Mavis Shezi) planted in line with CA principles. Below are a few snapshots
of the CA experimentation.
CA Strips
This season a new CA design was introduced which is planting in strips as an alternative to or in
conjunction with 10x10 plots, depending on the farmer. It was explained that the strips wouldbe
1000 m2insizeand wouldinclude both monocropand intercrop plots. The following farmers
planted CA strips:
Dumazile and Thembani Nxusa
Dumazile and Thembani Nxusa planted a2000m2strip in
November 2020 using the two-row tractor drawn planter,
whichincluded maize, maize and summer cover crops as well
as amaize and bean intercrop. They also planted a control plot
next tothe trial, which was ploughed and planted undera maize
monocrop. In terms of growth and cob sizes, there wasn’t a
significant difference between the two fields, except that the
trial plots seemed to have ahigher percentage of weeds. This
was in part due tothe farmersaccidentally killing off their beans
by using a broad leaf herbicide during the season- a practise not
promoted by the MDF team. This also meant the sunflowers did
not grow in their SCC plot.
There were alsogaps in the maize plantings, dueto damage by crows.Normally maize seed is
treated with Lavine in order to prevent crows from eatingit,
but this was not undertaken this season.
Figure 23: Right Above: A view of the strip cropped CA trial with Sun Hemp,
Sorghum visible onthe left and Maize (SC701) on the right. Right Below:The
layout of their strip cropping trial.
Fikelephi Maphumulo
She planted a 1000m2stripcropping trialof maize, beans and summer cover crops, with no
intercropped plot. She decided not to plough her field thisseason and thus her control plot is a
CA monocrop plot.
M
M+ SCC
M
M+B
Her reason for not ploughing was that since she started to incorporate CA principles, she noticed
over time that the erosion was muchless on her CA plot thanon the conventional plot. Also, the
crop yields from theconventional
plots seemed to be decliningwith
each season, and there were often
largepatches in between the maize.
Her other challenge is that her field is
on a steep slope and thus erosion and
runoffhave been amajor challenge
over the years. Lungile Madlala and
Nomusa Shandu, also have fields
situated on a steep slope and thus
have significant challenges with soil
erosion.
Figure 24: Right and Far Right: Fikile’s Maize control plot and her SCC strip showing the fodder sorghum in seed.
Lungile Madlala
She planted a 400m2
strip cropping CA trial
(SCC, M+Band M). In
general growth was
good, except for weed
competition in the
SCC plot. Growth was
somewhat patchy,
again due to damage
by crows.
Figure 25: Lungile’s Strip
cropping trial, planted on
contour across the steep slope of her field and growing well.
Mavis Shezi
Mavis Shezi is a 67 year old pensioner and very active farmer who resides with her husband. She
started planting CA trials in 2019 but had been attending meetings and farmers days prior out of
interest. Last season she planted a 400m2 plot of maize and beans and cowpeas and planted short
season maize. This season she planted a longstrip of 10 (10x5m)plots of M+B and M+SCC. She
has had challenges with stalkborer for which she sprayed with Decis forte (Cypermethrin), which
worked well. Stalk borer infestation of her adjacent control plot however, remained high. Overall
germination was very good and her crops were growing vigorously, despite patchy germination
in portions of the strips
Figure 26: Above Left: A strip of +=B where germinationwas patchy, here looking to be due to localised soil quality issues,
Above Centre: A strip of SCC growing very well and Above right: A view of Mrs Shezi’s top half of her field which is very steep
CA block trails (10x10m plots)
Babhekile Nene
Mrs Nene used the following layout.
The growth in different plots were quite dis-similar with some
being patchy, some growing well and one of the SCC plots, and one
of the bean mono cropped plotsshowing almost 0% germination.
This is in part due to expanding to a 1000m2 from her smaller plots
and now incorporating plots that had previously been unused or
planted to potatoes. Improved planting and crop management
would solve these issues.
Figure 27: in the foreground is a highly weed infested plot, where SCC did not
germinate, or were out competed. And in the background a maize and bean inter
cropped plot, showing reasonable growth.
Ntombi Shandu
Ntombi Shandu has been one of our most consistent and enthusiastic farmers and has been part
of the program since 2018. She also planted a block CA trial with the following layout:
Plot 1
Plot 2
Plot 3
Plot 4
Plot 5
Plot 6
Plot 7
Plot 8
Plot 9
Plot 10
M+B
M
B
SCC
M
B
M+B
M
M+B
M
Plot
1
Plot
2
Plot
3
Plot
4
Plot
5
Plot
6
Plot
7
Plot
8
Plot
9
Plot
10
M+B
M
B
SCC
M
M+B
M
B
M
SCC
She also hadthe issue that in expansion of her plot, she used the higher patches in her field, which
had been left fallow due to lack of productivity, in the last few years. The germination inthis
section was patchy and growth as moderate only. She planted winter cover crops in thespaces
where the maize did not germinate.
Figure 28: Above Left: Winter cover crops maturing in asection of the field where the maize and beans did not germinate
well. Above Centre: Poor germination higher up in Mrs Shandu’s field. And Above Right: A close up view fd winter cover crops
growing in between maize in a plot with patchy germination.
Nakeni Ngubane
She is also one of the mostconsistent farmers who is in her third season ofplanting under CA. She
decided to stick to the normal 400m2plot, possibly due to the limited area around her household,
as she uses the bottom section to plant other crops such as amadumbe, beans and vegetables. She
planted two maize and bean intercrop plots and a separate plot of the summer cover crop mix
and winter cover crops.
Figure 29: Above Left: A block of SCC with sunflowers dominating the mix. Above Centre: A M+B inter cropped plot maize is
a bit patchy and yellowing and Above right: A plot of winter cover crops growing well.
Yield measurements: Introduction
This season was impactedquite heavily by COVID-19 for the Midlands, which is closer to larger
cities and centres and saw much higher levels of infection than the other areas of implementation.
Three participating farmers passed away and a number of others became ill. Despite the
emotional and economic burden, most farmers participatedwell and the area saw a significant
increase in planting of cover crops.
Good rains in September and October meant that farmers could plant much earlier in the season
than past years and also impacted bean production positively. Below a small case study for
Gobizembe provides an indication of Ca production this season
Bean and CC yields
Higher humidity caused some difficulties for farmers inharvesting and causedsubstantial rotting,
both in intercropped and mono-cropped bean plots. One farmer attempted trellising thebeans,
but this had no significant impact on the yields. Bean yields were quite low, as can be seen in the
small table below. Very few participants managed to harvest seed from cover crops, these being
highly predated by both birds and buck in the area.
Table 17: Bean and CC yields for participants in Gobizembe 2021/21
The pictures below provide a snapshot of weighing for a number of participants in the Midlands
GOBIZEMBE BEAN YIELDS (2020/21)
CA Trial Bean Yield (from all plots)
SCC
Name
Surname
Experiment
Area (m2)
weight (kg)
t
t/ha
Crop name
weight (kg)
t/ha
Ntombiyomuntu
Ngobese
10x10
240
10,235
0,010
0,426
Janet Ntombencane
Gasa
10x10
200
9
0,009
0,450
Sorghum ,
mung beans
9and 1,2
0,9 and 0,1
Elijah
Ntuli
600m2
160
0
0,000
0,000
Sorghum
22,9
1,43
Khanyisile
Xasibe
400m2
160
2,5
0,003
0,156
Zwelinjani
Zuma
400m2
200
5
0,005
0,250
Sunflower
1,3
0,13
Khombisile
Mncanya
strips
200
14,15
0,014
0,708
400m2
160
15,523
0,016
0,970
Nelie
Ngcobo
100m2
50
5
0,005
1,000
Simephi
Choncho
200
5
0,005
0,250
TOTAL
0,468
Ntombi Shandu
Agnes Gabela
Babhekile Nene
Fikile Maphumulo
Nomusa
Shandu
Gogo Shangasee
Agnes Gabela
Fikelephi
Maphumulo
Lindeni
Nkala
Ntombi Shandu
Maize weighing
One interesting case, for Rita Ngobese (Gobizembe) where maize yields were separated by CA
plots for M only,M+B and M+SCC showed visible differences in the maize cob size and maize
yields for those plots.
Figure 30: Above Left to Right: Rita Ngobeses’s 10x10 trial maize separated for different plot types, shows bigger and better
filled cobs for M+B inter crops, than
Maize only and M+SCC, respectively.
For the latter, the cobs were much
smaller and incompletely filled,
indicating some competition between
maize and the SCC later in the season.
This effect has been noted previously
in Bergville as well.
In Gobizembe, the overall
performance of maize in this
season’s CA trials was very
poor, as most farmers either
got stunted and deformed
cobs or cobs went rotten
while in storage. The maize
this year seemingly had a very
high percentage of
mycotoxins, as most of the
maize in storage had fungal
growths barely a month after
it was harvested and left to
dry. The photos alongside are
indicative. Mycotoxin
samples have again been sent
to the ARC for analysis.
M+SCC
M
M+B
Ntombencane Gasa
Zwelinjani Zuma
Rita Ngobese
Elijah Ntuli
Southern KZN
SKZN bean yield summary and photographs
Beans are produced in both the block and strip cropping trial options either as single crops or as
intercrops with maize. Bean yields are generally not separated per plot by participants and thus
the bean yields recorded are the overall values for their CA experiments. In general, bean yields
have remained low for the entire experimentation period. Yield calculations take into account
whether beans were mono-cropped or inter-cropped.
Bean yields have been collectedinNgongonini, Springvalley, Madzikane,Plainhill, Nkoneni and
Ofafa. The small figure below provides a summary of yields ineach area. The average yield for
SKZN was 1,16ton/ha.
Figure 31: CA trial bean yields for 30 participants in SKZN.
Bean yields this season were average to low, followinga similar trend to previous years, but for
different reasons. This season heavy rains towards the end of the season damaged bean harvests
due to water damage and rotting of the pods and seeds.
Figure 32:Above Left: Leonard Gamede (Ngongonini), with his bean harvest, and Above Right: Bonginhlanhla Dlamini from
Spring Valley planted a late plot of beans, photographed here in May 2021, which showed promise of providing a good yield.
MadzikaneNgongoniniOfafaPlainhillSpring Valley
Total 1.08 1.32 1.10 0.72 1.20
0.00
0.20
0.40
0.60
0.80
1.00
1.20
1.40
ton/ha
CA bean yields SKZN 2020/21
Cover crops in SKZN
Very few farmers this season opted to plant cover crops. InSKZN grazing is not as limited as in
the other two areas and farmers are determined to focus solely on food crops in their farming.
Mthandeni Phungula (Spring Valley)
Hewas the only participant in SKZNto plant SCC. He intercropped the SCC with maize in one of
his plots. He planted millet and sunflower but lost his sunflower seed harvest to bird damage. His
millet yielded 18,2kg for
the100m2 intercrop.
He will be using themillet with
fermented maize forthe
brewing of traditional beer.
Maize yield considerations
Yield measurements wereundertaken for both the CA trial options where maize was either
planted as a single crop or as an intercrop with beans and cowpeas (M, M+B and M+C)and CA
control (maize only in consecutive years)plots. These trail options are the CMTs planted either
in the 10x10m block designs of 1000m2 or in the 1000m2 strip cropping trials.
For maize, the cob and grain weight for each participant was averaged prior to acount of the
number of cobs per50 kg bag and thena count of the number of bags per plot,to estimatethe
yields.
Bergville maize yields
In Bergville, yields for maize inthe CA trialand control plots were calculated fora selection of
participants (N=28) which had CMT’s. Yields were measured for 11 participants in Ezibomvini, 7
participants in Stulwane and 5 participants in Eqeleni and Vimbukhalo respectively.
Figure 33: Phungula's millet harvest,
maize seed kept for next season,
pumpkins from the intercrop and
maize soaked for imithombo
(Traditional beer).
Figure 34: Maize yields for CA trial and control plots for Bergville, 2020/21
From the above figure the maize yieldsfrom the CA trial plots are all higher and, in some cases,
(Eqeleni and Vimbukhalo), substantially higher than the maize yields for the CA control plots. The
difference in treatments between CAtrial plots and CA control plots is that the latter havebeen
planted to maize only throughout.
It is interesting to also note thatthe maize yields in the CA intercropped plots (here a combination
of M+B, M+C and M+Pk (Pumpkin)plots) are significantly higher than the maize yields in CA
mono-cropped plots.
Here the results indicate the potential for incremental improvement in maize yields using
the multi-cropping CA system,as opposed to aCA monocropping system where the yields
remained similar across six years of CA implementation.
Yields for the generic hybrid maize varieties PAN 6479 and PAN53 were comparable to the short
season variety (PAN5A190) planted later in the season in bothStulwane and Ezibomvini at
2,2t/ha and 6,3t/ha respectively. As in previous years, the favourable yield comparisons
indicate the planting of short season maize later in the season (January vs November) in
years with late onset rains and higher late season rainfall values as a good adaptive
strategy to compensate for changing weather patterns.
Table 18: Average yields for maize and beans in CA trial plots; Bergville (2014-2020)
Maize and bean yields (CA trial plots)
Bergville
Season
2014
2015
2016
2017
2018
2019
No. of villages
9
11
17
18
19
18
No. of trial participants
83
73
212
259
207
225
Area planted (trials) (ha)
7,2
5,9
13,5
17,4
15,2
13,9
Average yield maize (t/ha)
3,63
4,12
5,03
5,7
3,4
3,6
Min. and max. maize yield (t/ha)
1-6,7
0,6-7,4
0,3-11,7
0,5-12,2
0,1-8,5
0,5-12,8
Average trial quantity of maize (kg)
576
654
487
206
113
261
EqeleniEzibomviniStulwane Vimbukhalo
Average of yield(t/ha)CA-M8.35 7.29 2.21 6.67
Average of yield(t/ha)CA-M
intercrop 9.95 12.62 4.25
Average of yield(t/ha)CA9.21 7.90 2.96 6.67
Average of yield(t/ha)Control4.71 6.21 2.75 1.98
0.00
2.00
4.00
6.00
8.00
10.00
12.00
14.00
t/ha
Maize yields Bergville 2020/21(N=28)
Rand replacement value of maizemeal
R3 312
R4 120
R4 900
R2 350
R994
R1 850
WP for CA multi-cropping maize (kg/m3)
2,4
1,25
WP for CA mono-culture maize (kg/m3)
1,6
0,8
Average yield of beans (t/ha)
0,26
0,79
1,05
1,22
0,56
1
Due to climate variability (late onset rains and rainfall variability in season) the initial gains in
averagemaize yields underCA implementation from 3,6 t/ha to 5,7 t/ha on average between
2014 and 2017 could not be maintained into 2018 and 2019, but good average yields were again
attained in the 2020 season, which had high rainfall later in the season, despite late onset of rains.
Maximum yields obtained by individual smallholders have, however, increased from 6,7 t/ha to
14,8t/ha in that time, indicating that for high performing smallholder farmersa yield gain
of around 1 t/ha per annum is possible under CA cropping systems despite difficult
climatic conditions. During this period, average maize control plot yields did not increase
but remained stable at between 2,5 t/ha to 3,9 t/ha.
SKZN maize yields
Maize was weighed for 37 participants across five villages in SKZN, weighing CAtrial maize and
CA control maize for each participant.
The small graph below indicates the average yields of maize per village.
Figure 35: Maize yields for CA trial and control plots in SKZN 2020/21
In this region, the large majority of control plots were also CA plots but planted to mono cropped
maize only in consecutive years. Maize yields for the CA trial plots were higherthan the CA control
plots in Ngongonini andSpring Valley, butnot in the otherthree villages where yields were
measured. For Madzikaneand Ofafa the yields between the trial and control were very similar.
Overall, the CA trial maize yield was higher at 3,4t/ha than the CA control plot yield at 3,0t/ha.
In SKZN participants have been slow to take up the multi-cropping options and also slow
to rotate their intercropped plots. The yields of the CAtrial plots are thus comparable to
CA mono-cropped plots in this region.
MadzikaneNgongoniniOfafaPlainhillSpring Valley
Control 2.71 1.85 3.761.98
Trial 2.19 3.61 3.29 5.47 3.13
0.00
1.00
2.00
3.00
4.00
5.00
6.00
ton/ha
CA trial and control yields forSKZN 2020/21
A comparison of yields has been made for the period between 2014-2020/21.
Table 19: Average yields for maize and beans in CA trial plots: SKZN 2014-2020/21
Maize and bean yields (CA trial plots)
SKZN
Season
2014
2015
2016
2017
2018
2019
No. of villages
10
8
8
13
13
11
No. of trial participants
16
43
54
93
75
93
Area planted (trials) (ha)
0,3
0,37
1,18
3,58
4
3,72
Average yield maize (t/ha)
0,7
1,37
2,52
2,17
2,6
3,4
Min. and max. maize yield (t/ha)
0,3-1,8
0,5-4,4
1,1-5,2
0,2-6,7
0,2-6,9
0,3-9,6
Average trial quantity of maize (kg)
64
125
161
66
97
78
Rand replacement value of maizemeal
R500
R1 000
R1 700
R752
R854
R553
WP for CA multi-cropping maize (kg/m3)
1,73
WP for CA mono-culture maize (kg/m3)
0,9
Average yield of beans (t/ha)
1,26
0,34
0,69
1,28
0,35
0,6
From this table it can be seen that the average maize yield for 2020/21 is the same as the yield in
the previous season. As with Bergville the trend in increase in maximum yields for each season is
similar in SKZN where these have increased from1,8t/ha to 10,2t/ha in the observation period,
indicating that for high performing smallholder farmers an average yield gain of around 1
t/ha per annum is possible under CA croppingsystems despite difficult climatic conditions.
Midlands maize yields
Here yields werecompiled for different CA experimental protocols (10x10 blocks, strips and
control plots) for 50 participants from the 9 villages participating in the research process.
The small graph below indicates the average yields
Figure 36: Average yields for CA experimentation in the Midlands
In this region, the large majority of control plots were also CA plots but planted to mono cropped
maize only in consecutive years, with different inter row spacing and fertilization regimes to the
Ozwathini Swayimane
10x10 2.68 1.36
Control 2.57 1.51
Strips 4.49 1.86
0.00
1.00
2.00
3.00
4.00
5.00
t/ha
Maize yields for Midlands CA experimentation 2020/21
CA trial plots. Here yields for the strip cropping CA multi-cropping trials are higher than the CA
mono-cropped maize control plot sand the 10x10 block trial plots. Average yields overall in
Midlands for the CA trials is 2,6t/ha and for the CA controls is 2,0t/ha.
In the Midlands there is a slight yield advantage shown for strip cropping in CA versus the
10x10 blockstrialsthat havebeen used to date. Yields for the CA multi cropped and rotated
plots are higher than the CA mono cropped plots (controls).
Farmers in this area sell green maize to traders from thelarger city centres in the vicinity and
then keep some maize for drying and milling.
Incomes made from sales for green maize are shown in the table below.
Table 20: Incomes from green maize for the Midlands villages for 2020/21
Village name
Incomefrom Green
Maize
Total income/area
Average income/
area
Average
Income/m2
Ozwathini
Gobinsimbi
R1 890,00
Hlathikhulu
R1 200,00
Mbalenhle
R1 650,00
Mkhakhasini
R2 700,00
Swidi
R7 620,00
R15 060,00
R 1 225,00
R3,84
Swayimane
Gobizembe
R1 800,00
Mayizekanye 1
R17 965,00
Mayizekanye 2
R5 237,00
Mayizekanye 3
R59 807,00
R84 809,00
R 2 827,00
R7,40
Grand Total
R99 869,00
The maize yields for Ozwathini villages were almost double those in Swayimane, explaining the
averageincome/m2 for thesetwo areas at R7,40 and R3,84 respectively. Average income per
participant is around R2 000 from on average area of around 300m2 of maize. This would equate
to an income of around R66 600/ha. 80% of participants have soldsome of their maize and actual
incomes ranged from as little as R120 (400m2) to a maximum of R39 550 (2700m2)
This is a significantly higher income potential than selling dry maize per ton, which would equate
to around R10 400 for the 2,6t/ha average for these farmers.
Yields have been compared for the 2017-2020 planting season in the Midlands’
Table 21: CA trial yield comparisons for the Midlands for 2017-2020
Maize and bean yield for CA trial plots
Midlands
Season
2017
2018
2019
2020
No of villages
4
9
14
9
No of trial participants
42
70
164
87
Area planted (trials) - ha
1,36
3,1
4,2
4,3
Average yield CA maize (t/ha)
2,04
1,43
1,35
2,6
Min and max yield maize (t/ha)
0,4-7,1
0,3-4,6
0,5-6,3
0,3-5,8
Average trial quantity of maize (kg)
87
65
118
160
Rand value (maizemeal)
R992
R572
R836
R1 280
Average yield beans (t/ha)
0,62
0,87
0,78
1,1
From this table it can be seen that the average maize yield for 2020/21 is the slightly higher than
the yields in previous seasons. In this region there have beenno yield gains, but only a yield
stabilization for the CA plots. CA yields for the multi cropping options (inter cropping and crop
rotation) are around 23% higher than consecutively mono cropped maize.
Issues and recommendations
1.Despite the difficulties and inefficiencies resultant from the COVID-10 pandemic, most
smallholder farmers in the programme continued with their farmer led trials and worked
well with MDF staff to plan and monitor the CMT’s.
2.For the SKZN and Midlands areas, it will be important to be somewhat more directive in
terms of implementation of the trials, to be able to increase the multi-cropping and cover
crop planting in these areas, so that the experimentation can be comparable to that in the
Bergville region.
3.Smallholder farmers in SKZN have not shown an immediate interest in fodder production
and winter supplementation for livestock as a way of integrating livestock into their CA
systems. MDF willconsider finding at least one farmer in each area toundertake this
process and demonstrate the impacts to the rest of the CA learninggroups in the coming
season.
4.MDF is in theprocess of scouting forfunding to ensure fencing of the CAplots, as unfenced
fields are a major drawback restricting participants from meaningful contributions, both
in terms of cropping andsoil cover. The Okhahlamba LM have agreed toassist in the
Bergville area for 30 participants.
5.Collaboration withNWU and the ARC have provided valuable insights into soil health,
nematode population balances and mycotoxins and will be continued into the future.
References
A review of conservation agriculture research in South Africa. Swanepoel, Corrie M, Swanepoel,
Lourens H and Smith, Hendrik J. 2017. 2017, South African Journal of Plant and Soil, pp. 1-10.
FAO. 2003.Unlocking the Water Potential of Agriculture. [Online] 2003.
http://www.fao.org/3/y4525e/y4525e00.htm#Contents.
Giller, Ken E, et al. 2009.Review: Conservation agriculture and smallholder farming inAfrica;
The heretic's view. Field Crops Research. 114, 2009, pp. 23-34.
GRDC.2014.Measuring and monitoring changes in soil organic carbon accurately . s.l.:
Department of Agriculture and Food, Government of Western Australia, 2014.
Kruger, Erna and Smith, Hendrik, J. 2019.Conservation Agriculture Farmer Innovation
Programme (CA FIP) for smallholders, GrainSA. Farmer Centred Innovation inConservation
Agriculture in upper catchment areas of the Drakensberg, KwaZulu-Natal. Pretoria : GrainSA, 2019.
Long-term effects of no-tillage on soil physical properties in a Rhodic Ferrasol inParaná, Brzil.
Cavalieri, K M V, et al. 2009. 2009, Soil and Tillage Research, Vol. 103, pp. 158-164.
McKenzie, N J, et al. 2004.Australian Soilsand Landscapes: An Illustrated Compendium.
Collingwood, Victoria. : CSIRO Publishing, 2004.
Motaung, Palesa. 2020.Evaluating soil health of smallholder maize monocrops and intercrops
using qualitativeand quantitative soil quality assessment methods. Pretoria : MSc thesis in Soil
Science, University of Pretoria, 2020.
NRCS. 2011.Soil quality for environmental health: Bulk density. s.l. : NRCS and University of Illois,
USA., 2011.
Stolbovoy, Vladimir, et al. 2007.Soil sampling protocol to certify the changes of organic carbon
stock inminteral soil of the European Union. Version 2. Luxembourg. ISBN: 978-92-79-05379-5 :
EUR 21576 EN/2. 56 pp. Office for Official Publications of the European Communities, 2007.
APPENDIX 1: SAGL TEST REPORT NO (20/21) 27
SEASON:
2020/2021
DATE OF RECEIPT:
2020/10/07
ORIGIN:
ARC - Agricultural Research Council - Grain
Crops
SENDER:
Belinda Janse van Rensburg
ORDER NUMBER:
PSO-021485-1
PRODUCT:
Maize
LAB.
NO.:
SENDE
R'S
CODE:
Multi-Mycotoxin (In-House Method 26 (UPLC-MS/MS))
Aflatoxin ppb (µg/kg)
Fumonisin ppb
(µg/kg)
Deoxyniv
a-
lenol
(DON)
ppb
(µg/kg)
15-
Acetyl
-DON
ppb
(µg/k
g)
Ochra-
toxin A
ppb
(µg/kg)
Zearale-
none
ppb
(µg/kg)
T2
ppb
(µg/kg)
HT-2
ppb
(µg/k
g)
Diplodi
a-
toxin
ppb
(µg/kg)
B1
B2
G1
G2
Tot
al
B1
B2
B3
Limit of quantitation
(µg/kg)
5
5
5
5
-
20
20
20
100
100
5
20
20
20
50
ARC
code
Farmer information
Productio
n system /
rotation
2020/21
Ear
rot
ratin
g
(%)
(20/2
1) /
27 /
112
30
Babhekele Nene
(Mayizekane)
Monocultu
re maize
1
N
D
N
D
N
D
N
D
-
ND
ND
ND
ND
ND
ND
354
ND
ND
80
(20/2
1) /
27 /
114
31
Fikile Maphumolo
(Mayizekane)
3
N
D
N
D
N
D
N
D
-
ND
ND
ND
ND
ND
ND
976
ND
ND
ND
(20/2
1) /
27 /
115
32
Simephi Hlatswhayo
(Stulwane)
Monocultu
re maize
(plot 6)
0
N
D
N
D
N
D
N
D
-
ND
ND
ND
798
208
ND
144
ND
ND
ND
(20/2
1) /
27 /
116
33
Neliswe Msele
(Stulwane)
Maize-
cowpea
(plot 2)
0
N
D
N
D
N
D
N
D
-
ND
ND
ND
ND
ND
ND
29
ND
ND
ND
(20/2
1) /
27 /
117
34
Simephi Hlatswhayo
(Eqeleni)
Maize
(plot 4)
0
N
D
N
D
N
D
N
D
-
ND
ND
ND
240
<LOQ
ND
ND
ND
ND
ND
(20/2
1) /
27 /
118
35
Neliswe Msele
(Stulwane)
Maize
(plot 10)
0
N
D
N
D
N
D
N
D
-
ND
ND
ND
400
<LOQ
ND
134
ND
ND
ND
(20/2
1) /
27 /
119
36
Neliswe Msele
(Stulwane)
Maize
(plot 4)
0
N
D
N
D
N
D
N
D
-
ND
ND
ND
444
180
ND
78
ND
ND
ND
(20/2
1) /
27 /
120
37
Nomusa Shandu
(Mayizekane)
Maize-
cowpea
14
N
D
N
D
N
D
N
D
-
ND
ND
ND
ND
ND
ND
587
1409
1446
ND
(20/2
1) /
27 /
121
38
Neliswe Msele
(Stulwane)
Maize
(plot 8)
0
N
D
N
D
N
D
N
D
-
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
(20/2
1) /
27 /
122
39
Neliswe Msele
(Stulwane)
Maize
(plot 6)
0
N
D
N
D
N
D
N
D
-
466
92
57
1665
223
ND
49
ND
ND
ND
(20/2
1) /
27 /
123
40
Simephi Hlatswhayo
(Eqeleni)
Maize
(plot 8)
0
N
D
N
D
N
D
N
D
-
ND
ND
ND
530
168
ND
<LOQ
ND
ND
ND
(20/2
1) /
27 /
124
41
Fikile Maphumolo
(Mayizekane)
Monocultu
re maize
(ploughed
) (plot 3)
30
N
D
N
D
N
D
N
D
-
182
6
127
313
ND
ND
ND
21
ND
ND
ND
(20/2
1) /
27 /
125
42
Simephi Hlatswhayo
(Eqeleni)
Maize
monocultu
re (plot 2)
1
N
D
N
D
N
D
N
D
-
ND
ND
ND
613
144
ND
105
ND
ND
ND
(20/2
1) /
27 /
126
43
Phumelele Hlongwane
(Ezibomvini)
Maize
(plot 10)
0
N
D
N
D
N
D
N
D
-
104
53
ND
420
114
ND
<LOQ
ND
ND
ND
(20/2
1) /
27 /
127
44
Phumelele Hlongwane
(Ezibomvini)
Maize-
Beans
(plot 3)
0
N
D
N
D
N
D
N
D
-
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
(20/2
1) /
27 /
128
45
SKZN Madzikane
Pooled
maize
from
different
plots
0
N
D
N
D
N
D
N
D
-
ND
ND
ND
2050
316
ND
53
ND
ND
ND
(20/2
1) /
27 /
129
46
Babhekele Nene
(Mayizekane)
Maize-
Beans
5
N
D
N
D
N
D
N
D
-
556
111
85
ND
ND
ND
85
ND
ND
ND
(20/2
1) /
27 /
130
47
Nomusa Shandu
(Mayizekane)
Monocultu
re maize
12
N
D
N
D
N
D
N
D
-
471
9
123
3
944
ND
ND
ND
63
ND
ND
ND
(20/2
1) /
27 /
131
48
SKZN Madzikane
Monocultu
re maize
4
N
D
N
D
N
D
N
D
-
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
(20/2
1) /
27 /
132
49
Phumelele Hlongwane
(Ezibomvini)
Maize-
Beans
(plot 2)
0
N
D
N
D
N
D
N
D
-
424
106
65
373
127
ND
ND
ND
ND
ND
(20/2
1) /
27 /
133
50
Sibongile Mpulo
(Vimbukhalo)
Maize
(plot 4)
0
N
D
N
D
N
D
N
D
-
ND
ND
ND
ND
ND
ND
27
ND
ND
ND
(20/2
1) /
27 /
134
51
Letta Ngubo (SKZN
Springvalley)
Maize-
beans
0
N
D
N
D
N
D
N
D
-
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
(20/2
1) /
27 /
135
52
Sibongile Mpulo
(Vimbukhalo)
Maize
(plot 1)
0
N
D
N
D
N
D
N
D
-
23
ND
ND
ND
ND
ND
ND
ND
ND
ND
(20/2
1) /
27 /
136
53
Phumelele Hlongwane
(Ezibomvini)
Maize-
Beans
(plot 8)
0
N
D
N
D
N
D
N
D
-
116
27
ND
<LOQ
ND
ND
ND
ND
ND
ND
(20/2
1) /
27 /
137
54
Zweni Ndaba
(Emabunzini)
Maize-
Beans
(plot 1)
0
N
D
N
D
N
D
N
D
-
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
(20/2
1) /
27 /
138
55
Babhekele Nene
(Mayizekane)
Maize-
Cowpea
25
N
D
N
D
N
D
N
D
-
539
174
88
ND
ND
ND
4294
ND
ND
ND
(20/2
1) /
27 /
139
56
Nomusa Shandu
(Mayizekane)
Maize-
Beans
0
N
D
N
D
N
D
N
D
-
69
36
ND
849
105
ND
29
ND
ND
ND
(20/2
1) /
27 /
140
57
Phumelele Hlongwane
(Ezibomvini)
Maize
(plot 1)
1
N
D
N
D
N
D
N
D
-
157
77
ND
1950
177
ND
30
ND
ND
ND
(20/2
1) /
27 /
141
58
Phumelele Hlongwane
(Ezibomvini)
Maize-
Beans
(plot 7)
0
N
D
N
D
N
D
N
D
-
ND
ND
ND
379
121
ND
ND
ND
ND
ND
(20/2
1) /
27 /
142
59
Lethiwe Zimba
(Ndunwana)
Maize-
cowpea
0
N
D
N
D
N
D
N
D
-
ND
ND
ND
ND
ND
ND
56
ND
ND
ND
(20/2
1) /
27 /
143
60
Sibongile Mpulo
(Vimbukhalo)
Maize
(plot 10)
0
N
D
N
D
N
D
N
D
-
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
(20/2
1) /
27 /
144
61
Sibongile Mpulo
(Vimbukhalo)
Maize-
Cowpea
(plot 6)
0
N
D
N
D
N
D
N
D
-
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
(20/2
1) /
27 /
145
62
Phumelele Hlongwane
(Ezibomvini)
Maize-CC
(plot 4)
0
N
D
N
D
N
D
N
D
-
183
57
<LO
Q
1418
518
ND
107
ND
ND
ND
(20/2
1) /
27 /
146
63
Sibongile Mpulo
(Vimbukhalo)
Maize-
Beans
(plot 9)
0
N
D
N
D
N
D
N
D
-
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
(20/2
1) /
27 /
147
64
Letta Ngubo (SKZN
Springvalley)
Monocultu
re maize
0
N
D
N
D
N
D
N
D
-
81
<LO
Q
ND
ND
ND
ND
ND
ND
ND
ND
(20/2
1) /
27 /
148
65
Phumelele Hlongwane
(Ezibomvini)
Maize-
Beans
0
N
D
N
D
N
D
N
D
-
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
(20/2
1) /
27 /
149
66
Sibongile Mpulo
(Vimbukhalo)
Maize-
Beans
0
N
D
N
D
N
D
N
D
-
ND
ND
ND
399
<LOQ
ND
ND
ND
ND
ND