COMPARISON OF SOIL EROSION UNDER NO-TILL AND

COMPARISON OF SOIL EROSION UNDER NO-TILL AND
University of Pretoria – Kidson, M.V. (2014)
COMPARISON OF SOIL EROSION UNDER NO-TILL AND
CONVENTIONAL TILLAGE SYSTEMS IN THE HIGH RAINFALL
MLONDOZI AREA, MPUMALANGA PROVINCE, SOUTH AFRICA
MICHAEL VERNON KIDSON
Submitted In partial fulfilment of the requirements for the degree M.
Inst Agrar (Land-use Planning)
In the Faculty of Natural and Agricultural Sciences
University of Pretoria
Department of Crop Production and Soil Science
Faculty of Natural and Agricultural Sciences
University of Pretoria
2014
I
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
DECLARATION
I hereby certify that this thesis is my own work, except where duly acknowledged. I also certify that
no plagiarism was committed in writing
...........................................................................................………………
Michael Vernon Kidson
31-03-2014
II
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
ACKNOWLEDGEMENTS
The Mlondozi Dryland Maize project that developed into the Mlondozi LandCare project was a project
that was aimed at developing the agricultural practices of the people in Mlondozi. It was the first
LandCare project in South Africa. We learnt a lot, but more importantly we achieved a lot of successes.
Special thanks are due to Hester Jansen van Rensburg, project leader, scientist, leader and motivator
extraordinaire.
This was a project for a community. Due to the complexities of the project, a number of experts were
included in the project team. This included the farmers in Mlondozi, the Mpumalanga Department of
Agriculture, especially Simon Tshabalala for his tireless enthusiasm and energy; Dr Jim Findley who
taught us to teach the communities; Johann Adendorff, an inspiration and pioneer in community
development; our sister Institutes the Plant Protection Research Institute, Jacomina Bloem and Dr
Jolisa Pakela, the Institute for Agricultural Engineering Johan Fuls, Johan van Biljon and Gawie Stols;
Animal Production Institute Gerry Trytsman, as well as Marie Smith for her much valued contribution
with the statistical analysis.
A special thank you to Nicolene Thiebaut for assistance with statistical analysis and the re-write; also
Prof. Robin Barnard who has been my supervisor.
To the host of unmentioned people, who contributed to the development of the Mlondozi community,
thank-you.
To my family who have been my inspiration and untiring support.
Michael Kidson
III
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
ABSTRACT
Comparison of soil erosion under no-till and conventional tillage systems in the high
rainfall Mlondozi area, Mpumalanga province, South Africa
By
Michael Vernon Kidson
Study leader:
Prof R. Barnard
Department:
Department of Crop Production and Soil Science
Degree:
M. Inst Agrar (Land-use planning)
Rural agriculture in Mlondozi, as for South Africa, is has a low productivity, which is the result of poor
knowledge, information, beliefs and land tenure which limits the acquiring of loans for inputs.
A LandCare Project was conducted for four years. Training was in the form of farmer managed research
demonstrations which included the taught Conservation agriculture farming system which they
compared to the Traditional farming system. Eighteen farmers initially joined the program and their soils
were monitored for four years. At the end of the project undisturbed soil samples were taken from their
fields where maize was cultivated following no-till (NT) farming system, and the conventional tillage
system (CT).
The soils from the two farming systems were compared using a laboratory rainfall simulator for run-off,
erosion and infiltration. Each storm event in the rainfall simulator lasted for a period of 110 minutes (50
rotations). There were two statistical analyses done on the results. The first was a t-test was applied to
the data to test for differences between the two systems, with a sample size of 72, at 18 sights with 4
replicates, except for carbon which was 36 analyses for the 18 sights. There was a significantly higher
soil loss for NT soils for storm 1, compared to the CT soil, and a non significant difference for run-off for
IV
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
storms 1 and 2. Infiltration was significantly lower for the NT soil for the first storm, and not significantly
higher for the NT soil for the second storm.
The CT soils had a significantly higher infiltration rate for the first 16 rotations. After 68 minutes (rotation
34) the NT soils infiltration rate was higher. For simulated storm 2 the CT soils had a slightly higher
infiltration rate up to 32 minutes (16 rotations) where after NT soils had a higher infiltration rate. Between
48 and 80 minutes (rotations 24 and 40) the NT soils had a significantly higher infiltration rate. From the
results it can be concluded that the NT soils maintained the aggregate stability far longer than the CT
soils.
The initial and final infiltration rates were compared for the NT and CT soils for the simulated storms 1
and 2. The CT soils’ initial and final infiltration rate was similar, while the NT soils had a higher initial
and final infiltration rate for the second storm, due to the soils settling with the first storm. The results
question current literature that states that sealing of soils is a permanent feature. The carbon content
of the NT soils was not significantly higher than the CT soils, which corresponded with the results.
V
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
DEDICATION
To God my saviour,
My father who led by example
VI
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
TABLE OF CONTENTS
CHAPTER 1………………………………………………………………………………….
1
INTRODUCTION…………………………………………………………………………
1
1.1 GENERAL BACKGROUND……………………………………………………….
2
1.2 THE MLONDOZI LANDCARE PROJECT……………………………………….
3
1.3 LOCATION AND GENERAL DESCRIPTION OF THE STUDY AREA……….
3
1.4 PHYSICAL AND BIOLOGICAL RESOURCES OF MLONDOZI………………
5
1.4.1 Climate………………………………………………………………………….
5
1.4.2 Geology…………………………………………………………………………
11
1.4.3 Soils……………………………………………………………………………..
11
1.5 VEGETATION……………………………………………………………………….
16
1.6 INTERVENTION AND PRODUCTION SYSTEMS PRACTICED IN MLONDOZI.. 16
1.7 OBJECTIVES OF THE STUDY……………………………………………………
17
CHAPTER 2………………………………………………………………………………….
18
LITERATURE STUDY……………………………………………………………………
18
2.1 A BRIEF OVERVIEW OF AGRICULTURE IN RURAL AREAS………………..
18
2.2 CONSERVATION AGRICULTURE……………………………………………….
22
2.3 FACTORS AFFECTING SOIL STABILITY AND INFILTRATION……………..
26
2.3.1 Crust formation…………………………………………………………………
27
2.3.2 Parent material…………………………………………………………………
28
2.3.3 Degree of weathering…………………………………………………………
29
2.3.4 Free iron and aluminium oxides……………………………………………..
29
2.3.5 Clay mineralogy……………………………………………………………….
30
2.3.6 Sodium and magnesium……………………………………………………..
30
2.3.7 Organic material………………………………………………………………
31
2.3.8 Slope……………………………………………………………………………
32
VII
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
2.4 METHODS OF SOIL EROSION RESEARCH……………………………………
33
CHAPTER 3…………………………………………………………………………………..
36
MATERIALS AND METHODS…………………………………………………………..
36
3.1 SITE SELECTION…………………………………………………………………..
36
3.2 PRODUCTION SYSTEMS PRACTICED IN MLONDOZI………………………
38
3.2.1 Conventional Cultivation (CT) production system…………………………..
38
3.2.2 No-Till production system……………………………………………………..
42
3.3 CLASSIFICATION AND PROPERTIES OF THE SOILS OF THE STUDY SITES. 42
3.4 LAORATORY RAINFLL SIMULATOR…………………………………………….
42
3.5 SOIL SAMPLING FOR THE RAINFALL SIMULATOR…………………………..
44
3.5.1 Sampling equipment……………………………………………………………
45
3.5.2 Sampling procedure…………………………………………………………….
46
3.5.3 Transferring the soil samples to the plastic trays ……………………………
47
3.6 YIELD ESTIMATION…………………………………………………………………
48
3.7 THE LABORATORY RAINFALL SIMULATOR……………………………………
48
3.8 DETERMINATION OF IN-FIELD MULCH COVER……………………………….
49
3.9 DETERMINATION OF THE IN-FIELD ROOT COUNT…………………………..
50
3.10 OTHER LABORATORY ANALYSIS……………………………………………….
50
3.11 STATISTICAL ANALYSIS………………………………………………………….
51
3.11.1 Fertility……………………………………………………………………………
51
3.11.2 Maize yield……………………………………………………………………….
52
3.11.3 Mulch cover and Root count……………………………………………………
52
3.11.4 Runoff, Infiltration and Soil loss………………………………………………..
52
3.11.4.1 Statistical Analysis part I…………………………………………………
52
3.11.4.2 Statistical Analysis part II………………………………………………..
53
CHAPTER 4……………………………………………………………………………………
54
RESULTS AND DISCUSSION…………………………………………………………..
54
4.1 SOIL CLASSIFICATION AND MINERALOGY
54
……………………………
VIII
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
4.2 SOIL SAMPLING USING THE PAN METHOD..…………………………………
55
4.3. RESULTS AND STATISTICAL ANALYSIS…………………………..................
56
4.3.1 Fertility……………………………………………………………………………….
56
4.3.2 The Comparison of NT and TC farming systems – Yield………………………
58
4.3.3 Mulch cover and Root count……………………………………………………
62
4.3.4 Runoff, Infiltration and Soil loss for the NT and CT treatments…………….
64
4.3.4.1 Statistical Analysis - part I……………………………………………..
64
4.3.4.2 Statistical Analysis - part 2…………………………………………….
70
CHAPTER 5……………………………………………………………………………………
75
CONCLUSIONS AND RECOMMENDATIONS FOR FURTHER STUDY ………………
75
REFERENCES…………………………………………………………………………………
80
IX
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
FIGURES
Figure 1.1 Orientation maps, showing Mlondozi located on the eastern
border of Mpumalanga, and the western border of Swaziland……………...
4
Figure 1.2 Long term monthly average, temperature and rainfall for Mlondozi ……..
7
Figure 1.3 Long term average monthly relative humidity for Mlondozi (Athole)……..
7
Figure 1.4 Broad rainfall patterns of South Africa. ……………………………………..
8
Figure 1.5 Aridity zones of South Africa. ……………………..………………………..
9
Figure 1.6 Rainfall erosivity map of South Africa ………………………………………
10
Figure 1.7 Soil map of Mlondozi (Booyens, Potgieter and Matlawa 2000)………….
13
Figure 3.1 Mlondozi study area showing the distribution of the study sites…………
37
Figure 3.2 Metal tray used to take samples in field.……………………………………
44
Figure 3.3 Soil sample transferred to the plastic tray………………………………….
47
Figure 3.4 Example of a perforated plastic tray ……………………………………….
48
Figure 3.5 Schematic diagram of the main parts of the laboratory rainfall simulator…
51
Figure 4.1 Average of the Ca, Mg, of the NT soils for the 18 study sites………………
57
Figure 4.2 Average of the K and P of the NT soils for the 18 study sites………………
57
Figure 4.3 Average of the CEC of the soils in the NT fields of the 18 study sites……
57
Figure 4.4 Average of the NT and CT yields for the study period………………………
62
Figure 4.5 Graph of the mean soil loss ………..............................................................
64
Figure 4.6 Graph of the mean run-off ..……………………………………………………
65
Figure 4.7 Graph of total infiltration for the first and second simulated storm event …
65
Figure 4.8 Infiltration means in the NT and CT for storm 1…..….……….………………
68
Figure 4.9 Infiltration means in the NT and CT soils for storm 2………………………..
68
Figure 4.10 Average infiltration rate for the 2 storms for each rotation…………………
73
Figure 4.11 Average infiltration means as per treatment (NT and CT). …………………
74
X
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
TABLES
Table 1.1 Table of long term data for Athole and Oshoek weather stations....................... 6
Table 1.2 Extended legend for the soil map of Mlondozi……...……………………………..
14
Table 2.1 Nutrient composition of dairy manure……….……………………………………… 19
Table 2.2 New threshold slope percentage for the arable soils of Mavuso pedosystems… 33
Table 3.1 Names of farmers and co-ordinates for the 18 study sites ……..……………….. 38
Table 4.1 Soil classification………………………………………………………………………. 54
Table 4.2 Clay mineralogical analysis of the soils from the 18 study sites ………………… 55
Table 4.3 Combined ANOVA over seasons….………………………………………………… 58
Table 4.4 Yields of the NT and CT field of the 18 farmers..………………….………………. 59
Table 4.5 ANOVA table for yields……………………………………………………………….. 60
Table 4.6 Means of yield for the two treatments …..………………………………………….. 61
Table 4.7 Mean yield over the different years…….…………………………………………… 61
Table 4.8 Mean yield for different treatments over the seasons……………………….......... 61
Table 4.9 The average mulch of the NT plots over the study period …..………………….. 62
Table 4.10 Average root counts for the NT fields……………………………………………… 63
Table 4.11 Mean soil loss, run-off and total infiltration (n=72)……………………………….. 64
Table 4.12 Average infiltration means for the eighteen NT and CT soils (n=72)…………... 67
Table 4.13 Average initial (IIR) and final infiltration rates (FIR) ……………………………… 69
Table 4.14 ANOVA for the farming systems NT and CT (combined)…………………........ 71
Table 4.15 Average infiltration means for the NT and CT soils and different rotations…… 71
Table 4.16 Average infiltration rate for the 2 storms combined as per treatment…………. 73
XI
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
ABBREVIATIONS
ARC
Agricultural Research Council
CA
Conservation Agriculture:
CT
Conventional tillage
CO2
Carbon dioxide
EO
Extension Officer
FIR
Final Infiltration Rate
IIR
Initial infiltration rate
ISCW
Institute for Soil, Climate and Water
KCl
Potassium chloride
LAN
Limestone ammonium nitrate
MDACE
Mpumalanga Department of Agriculture, Conservation and Environment
MDC
Mpumalanga Development Corporation
MLPC
Mpumalanga Liming Project Committee
MDML
Mlondozi Dryland Maize liming project
NDA
National Department of Agriculture
NGO
Non-Government Organisation
NLP
National LandCare program
NT
No Till
PDA
Participatory Development Approach
PPRI
Plant Protection Research Institute
SOC
Soil organic carbon
SOM
Soil organic material
XII
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
CHAPTER 1
INTRODUCTION
This study investigates soil erosion under two tillage systems: No-Till and Conventional tillage.
The two systems differ in a number of respects, all of which contribute to some extent to the
relative disturbance of the soil that occurs.
The “Conventional tillage” (CT), included:
•
typical low input;
•
ploughing of the soil;
•
grazing the crop residue and minimum weed and pest control;
as typically practised in traditional rural farming systems.
The “No-Till” (NT), curtailed soil disturbance, on the other hand, included a number of
additional enhancements, normally associated with Conservation Agriculture (CA), such as:
•
Minimum soil disturbance;
•
Keeping the soil covered with living or dead mulch;
•
Crop rotation;
•
Integrated soil fertility, weed control and pest and disease control.
The LandCare programme, adopted by the Department of Agriculture in 1995, is an activity
that gets the community actively involved in improving the delivery and adoption of soil
conservation practices. It had its origins in Australia, where it became a national programme
in 1992. One of the main objectives of the LandCare Program was to demonstrate the
differences between the two tillage systems in order to achieve a paradigm shift with the
farmers. Any differences in yield and chemical characteristics experienced cannot be
1
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
explained in terms of soil tillage, per se, but rather to the differences in inputs and management
between the two systems.
1.1
GENERAL BACKGROUND
Mlondozi was one of three blocks in Mpumalanga, which made up the district of KaNgwane.
This so-called homeland was formed in 1982, and on 27 April 1994 the territory was
reincorporated into South Africa (KaNgwane, 2010).
Mlondozi is located in a high rainfall
area and contains many highly weathered and highly leached soils. Consequently high soil
acidity is a major problem. Over 90% of the soils were found to have a pH KCl below 4.2.
The arable soils have a high potential for good yields, but the soils are not delivering the full
production potential due to poor farming practices.
The Mlondozi district has a Land
tenureship system, which means that the land is distributed by the chiefs, so that the farmers
do not own fields of their own. There exists the risk of the field being taken away by the chief,
especially if the field is not planted for a particular season.
Because the farmers do not have ownership of the lands they are not able to get production
loans for the purchasing of inputs such as lime and fertilizer (D’Haese and Kirsten, 2006). The
insecurity of tenure has further resulted in little incentive for soil conservation (Xaba, 2002).
The final result is that the average maize yield is very low - at only 0,5 ton ha-1, whereas the
potential was 4,5 ton ha-1. With the circumstances above, especially the generally good
rainfall, there exists a need to improve the yield without excessive cost and effort. The
improved farming system would lead to the protection of the natural resources and improved
productivity.
2
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
1.2
THE MLONDOZI LANDCARE PROJECT
The Agricultural Research Council – Institute for Soil, Climate and Water (ARC-ISCW) initiated
the Mlondozi LandCare project under the SoilCare theme of the National Landcare
Programme (NLP). The project was conducted in collaboration with the Southern Highveld
Region Extension, Mpumalanga Department of Agriculture, Conservation and Environment
(MDACE), Mlondozi farmers and the farmers’ associations.
The goal of the Mlondozi
LandCare project was to demonstrate and assess sound land management practices, by
involving local communities, who would contribute to sustainable and profitable agricultural
production in the district.
As the main thrust of this project “No-Till” (Conservation Agriculture (CA)), was introduced with
the objective of improving yields, following a sustainable farming system with low input costs.
This was compared to conventional tillage system typically used in the area.
In this thesis the two farming practices, “No-Till” and “Conventional Tillage”, were compared,
mainly for soil erosion effects.
1.3
LOCATION AND GENERAL DESCRIPTION OF THE STUDY AREA
Mlondozi is situated in the eastern Highveld of Mpumalanga, bordering Swaziland, as
indicated in Figure 1.1. The borders are formed by the Lochiel - Oshoek road (N17) while the
Amsterdam municipal district borders the area on the western and southern sides. It is situated
between 26o 05’ and 26o 30’ S and 30o 44’ and 31o 00’ E.
Mlondozi occupies an area of approximately 54 000 ha. About 3 553 ha (7%) of Mlondozi is
occupied by villages and 13 497 ha (25%) is under exotic timber plantations.
It has
approximately 12 746 ha (23.6%) potentially arable land, including 5 619 ha (10.4%) high
3
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
potential arable land. The proportions of total arable and high potential arable land are much
higher than the respective averages for South Africa, viz. 13% for total arable land and 3% for
high potential arable land. An estimated 4000 ha of the potentially arable land is currently
being cultivated. Maize is the dominant crop, while there are a number of vegetable gardens,
often under supplemental irrigation, largely managed by women’s clubs. The crops are very
important to livelihoods, providing the basic food requirements of the people (Jansen van
Rensburg, 2009).
Figure 1.1
Orientation maps, showing Mlondozi on the eastern border of
Mpumalanga, and on the western border of Swaziland. Included are the farmers
4
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
1.4
PHYSICAL AND BIOLOGICAL RESOURCES OF MLONDOZI
1.4.1
Climate
Mlondozi has a large fluctuation of seasonal, monthly as well as day and night temperatures.
The monthly average daily temperatures range from 10.4 °C for the coldest month (June) to
18.3 oC for the hottest months (December to February), with a mean frost season length of 64
days (Agromet, 2010). Long term rainfall for the district is 975 mm in the West and 892 mm
in the East as indicated in Table 1.2 and Figure 1.3. Mlondozi is therefore regarded as a high
rainfall region (Figure 1.4) with 800 mm plus per annum. Mlondozi has a high humidity (Figure
1.5, Agromet 2010), due to the high rainfall, low evapotranspiration and mist rain. Although
Mlondozi experiences a high rainfall it has a low rainfall erosivity index (Figure 1.6).
The ARC – Institute for Soil Climate and Water has 570 automatic weather and 80 mechanical
weather stations distributed across the country, giving relatively wide coverage of climatic
events. The average long term data for Mlondozi district was measured at Oshoek (19652002) and Athole (1936-2004), representing the North and South respectively. This selection
was made due to the different climatic zones (Table 1.1, Fig.1.2 & 1.3). The Athole weather
station is located 10 km from the study site on the South Western border, and the Oshoek
weather station, at the North Eastern corner of the study site, with the following location
information:
Athole weather station:
Coordinates:
Latitude -26.60;
Altitude:
1346 m
Longitude 30.58
Weather station number: 16301
Oshoek weather station:
Coordinates:
Latitude -26.22;
Altitude:
1470 m
Longitude 30.98
Weather station number: 16486
5
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
Table 1.1 Table of long term data for Athole (1936-2004) and Oshoek (1965-2002) weather stations (Agromet, 2010).
Elem
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Dec
Tx1
Tn1
Tave1
RHx1
RHn1
RHave1
Rain1
Rain2
23.8
12.9
18.4
86.8
40.3
63.5
161.2
147.33
23.6
13.0
18.3
88.8
42.0
65.4
142.6
118.26
23.0
12.0
17.5
88.0
38.7
63.3
97.3
105.19
21.6
9.5
15.6
82.0
31.6
56.8
50.9
50.09
19.5
6.3
12.9
71.9
26.1
49.0
16.7
35.1
17.2
3.4
10.3
65.5
23.7
44.6
15.2
10.04
17.6
3.2
10.4
66.9
22.5
44.7
10.2
8.89
19.9
5.1
12.5
69.5
23.9
46.7
16.9
14.78
22.4
8.0
15.2
70.9
25.6
48.2
44.5
36.58
22.9
10.0
16.4
76.8
31.7
54.2
105.5
99.71
23.2
11.5
17.3
80.8
36.2
58.5
142.4
138.56
24.0
12.6
18.3
84.9
38.1
61.5
171.8
127.34
Rain1: Athole weather station, 10 km from the study site on the western border (mm)
Coordinates:
Latitude -26.60; Longitude 30.58
Altitude:
1346 m
Weather station number: 16301
1
Tx :
Average Daily Maximum Temperature (oC)
Tn1:
Average Daily Minimum Temperature (oC)
Tave1: Average temperature (oC)
RHx1: Average daily maximum relative humidity (Ratio)
RHn1: Average daily minimum relative humidity (Ratio)
RHave1: Relative humidity average (Ratio)
Rain2: Oshoek weather station, at the north eastern corner of the study site (mm)
Coordinates:
Latitude -26.22; Longitude 30.98
Altitude:
1470 m
Weather station number: 16486
6
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
200.0
180.0
160.0
140.0
120.0
100.0
80.0
60.0
40.0
20.0
0.0
20.0
18.0
16.0
14.0
12.0
10.0
8.0
6.0
4.0
2.0
0.0
Temperature (oC)
Rainfall (mm)
Longterm rainfall and temperature data for Mlondozi
Rain2
Rain1
Tave
Jan Feb Mch Apl May Jun Jul Aug Sep Oct Nov Dec
Months
Figure 1.2 Long term average monthly, temperature and rainfall for Mlondozi
(Athole(Rain1): 1936-2004) and (Oshoek(Rain2): 1965-2002)
Long term monthly average relative humidity
70.0
60.0
Percentage
50.0
40.0
RHave
30.0
20.0
10.0
0.0
Jan Feb Mch Apl May Jun
Jul
Aug Sep Oct Nov Dec
Figure 1.3 Long term average monthly relative humidity for Mlondozi (Athole; 19362004 only)
7
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
Figure 1.4 Broad rainfall patterns of South Africa. (ARC-ISCW, 2004)
8
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
Figure 1.5 Aridity zones of South Africa. (ARC-ISCW, 2004)
9
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
Figure 1.6 Rainfall erosivity map of South Africa (ARC-ISCW, 2004)
10
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
1.4.2
Geology
Granites occurring in Southern Mpumalanga and KwaZulu-Natal are located in three zones,
and are of Swazian, Swazian to Randian and Namibian ages.
The mapping unit is
described as the Transvaal belt of granitization and metamorphism (Geological Survey,
1970 as cited by Turner, 2004). In a subsequent edition (Geological Survey, 1984) it is
described as a number of units including the Mpulusi Granites in the south. It comprises
unnamed potassic granite (light coloured granite) along with granodiorite (biotite granite)
labelled ZB (Kent, 1980; Alberts, 1986, as cited by Turner, 2004). Dolerite intrusions are
common in the district (Personal communication, Chris de Jager, 2010).
1.4.3
Soils
Prior to the start of the Mlondozi LandCare project, the only available soil data for the region
were the data from the national land type survey at 1:250 000 scale. During the project a
semi-detailed survey was conducted (Booyens, Potgieter & Matlawa, 2000). Observations
were made at 500 m intervals, using a fixed grid system, for areas which were intensively
cultivated, amounting to one observation per 25 ha. Plantations and very rocky areas were
surveyed with observations at intervals of up to 1 000 m. The soil map units consisted of
soil associations. The soil map was published at a scale of 1:50 000. Due the low density
of observations, even in the most intensively surveyed areas, the nature of the composition
of map units and the medium scale of the map, the map is only suitable for the identification
of promising areas for crop production. It cannot be used for farm planning, especially for
small-scale agriculture.
The soils were classified according to Soil Classification: A Taxonomic System for South
Africa (Soil Classification Working Group, 1991). The dominant soils in each map unit were
11
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
classified at family level and sub-dominant soils at form level. Soils of the Mispah form
(shallow soils on hard rock) and rock outcrops dominate most of the surveyed area
(Booyens, Potgieter & Matlawa, 2000). Patches of deep, high quality soils of the Clovelly
form are found throughout these areas. It would be important to identify these by means of
detailed soil surveys, where applicable.
The Clovelly and Magwa soil forms (Figure 1.7 & Table 1.2) are the most prominent
(dominant) soils found in the surveyed area. The soils reflect the influence of the two major
soil forming factors, viz. climate and geology. Magwa and Inanda soils have humic A
horizons, i.e. topsoil horizons with high organic matter contents and low base status.
Clovelly and Hutton soils of the area are dystrophic families of these two forms, i.e. highly
weathered, highly leached soils with a low base status. This means that all these soils have
low fertility and are strongly acidic. The high rainfall has led to intensive weathering and
leaching, while the moderate temperatures and high humidity produce and preserve high
levels of organic matter. The predominantly granite parent material (a felsic or “acid”
igneous rock) contributed further to the situation mentioned above.
The topography of the area led to the development of the sub-dominant soils, viz. shallow
soils (Mispah, Glenrosa) on the top slopes, the dominant Magwa/ Clovelly /Inanda/ Hutton
soils on the middle slopes and soils with sub-soils showing different degrees of wetness
(Avalon, Pinedene, Longlands, Kroonstad, Katspruit) on the foot slopes.
12
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
Figure 1.7 Soil map of Mlondozi (Booyens, Potgieter and Matlawa 2000).
13
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
Table 1.2 Extended legend for the soil map of Mlondozi, Figure 1.7 (Booyens, Potgieter & Matlawa, 2000)
Dominant soil
Dominant
form
depth (m)
Magwa
Clovelly
Other soil forms
Description
Mispah, Glenrosa
Shallow to moderately deep, strong brown to dark yellowish brown,
0.3 – 0.6
structureless, sandy clay loam soil on highly weathered rock and solid
rock. Deeper soil forms occur. Rock outcrops comprise between 0-15%
of mapping unit.
Magwa
Clovelly
0.6 – 0.9
Avalon, Kroonstad
Moderately deep, strong brown to dark yellowish brown, structureless,
Katspruit, Mispah,
sandy clay loam soil on highly weathered rock and solid rock. Rock
Longlands, Glencoe,
outcrops comprise between 0-15% of mapping unit
Pinedene, Oakleaf,
Glenrosa
Magwa, Inanda
Clovelly, Hutton
0.6 – 0.9
Magwa
Clovelly
0.6 – 0.9
Kroonstad, Avalon,
Moderately deep, strong brown to dark yellowish brown to yellowish red,
Mispah, Longlands,
structureless, sandy clay loam soil on highly weathered rock and solid
Oakleaf, Glenrosa
rock. Rock outcrops comprise between 0-15% of mapping unit
Avalon, Kroonstad,
Moderately deep soils. Rock outcrops comprise between 0-15% of
Katspruit, Longlands,
mapping unit
Hutton, Inanda
Magwa
Clovelly
> 0.9
Avalon, Kroonstad,
Deep, strong brown to dark yellowish brown, structureless, sandy clay
Katspruit, Mispah,
loam soil on highly weathered rock and weathered rock. Shallower soil
Glenrosa
forms do occur. Rock outcrops comprise between 0-15% of mapping
unit
14
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
Table 1.2 Extended legend for the soil map of Mlondozi, Figure 1.7 (Booyens, Potgieter & Matlawa, 2000)
continued
Dominant soil
Dominant
form
depth (m)
Magwa
Clovelly
> 0.9
Other soil forms
Description
Avalon, Inanda, Hutton,
Deep structureless soils. Rock outcrops comprise between 0-15% of
Kroonstad, Katspruit,
mapping unit
Mispah
Magwa
Clovelly
0.3-1.5
Avalon, Inanda, Hutton,
Structureless soils with variable depth. Rock outcrops comprise
Kroonstad, Katspruit,
between 20-30% of mapping unit.
Mispah
Mispah
<0.3
Clovelly
Rock outcrops dominant. Some deep soils do occur.
15
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
1.5
VEGETATION
The district of Mlondozi is situated in a Grassland Biome (Rutherford & Westfall, 1986).
Acocks (1988) classified the veld as Veld Types 63 or Piet Retief Sourveld with a small
intrusion of veld type 57 or North-eastern Sandy Highveld. The reason for the inclusion as
a false grass-veld type is because there are indications that the area was previously
Bushveld or Thornveld of an open sour type. The condition of the veld was regarded as
good (Myburg & Breytenbach, 2001).
1.6
INTERVENTION AND TRADITIONAL PRODUCTION SYSTEM PRACTISED IN
MLONDOZI
In the Mlondozi area the traditional farming system followed generally gave low maize
yields. An average yield of 0,4 ton per ha was determined at the inception of the project.
The practice consisted of relatively low inputs, no liming although the soils were acidic,
shallow ploughing, limited weed control, late planting and generally low levels of
management.
With the current intervention, it was decided to compare the conventional tillage (CT) with
the “No Till” (NT) approach, using realistic fertilization, liming, herbicides, pesticides, hybrid
seed and appropriate planting dates.
The “No Till” (NT) farming system consisted of farmer led trials, involving 18 farmers who
were taught how to implement this farming system. This system is a basket of technologies
starting with three main components, namely:
•
Minimum soil disturbance
•
Keeping the soil covered with a mulch
• Crop rotation
16
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
This was augmented by adopting integrated fertility, and integrated weed and pest
management.
The farmers’ training followed the 80/20 principle. The 20% consisted of a short theoretical
training session on conservation agriculture and the implementation thereof. This was
followed by support in the practical implementation of the new farming system, starting with
liming of the fields in winter, followed by spraying of weeds in spring. The necessary
resources were given to each of the 18 farmers, to plant a quarter hectare, with maize,
following the NT farming and CT systems with the layout as a spilt-plot design. The logic
behind supplying inputs to the farmers was two-fold:
1) To reduce the risk to the farmer while implementing the new technology and
2) For the farmer to produce sufficient maize for household consumption and selling the
remaining maize so as to purchase more inputs to expand “No Till” to the rest of his/her
farm.
In 1997 the Mlondozi Liming Project started, where lime was distributed to registered
farmers. This was followed in 1999 by the LandCare project, where the “No Till” (NT) and
CT were compared.
1.7
OBJECTIVES OF THE STUDY
•
Determine the differences between the NT and CT treatments for infiltration, run-off
and soil loss, at the end of a five year implementation period, by using a laboratory
rainfall simulator.
•
Compare the treatments for biological measurables such as fertility, root counts,
yields and mulch cover.
17
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
CHAPTER 2
LITERATURE STUDY
2.1
A BRIEF OVERVIEW OF AGRICULTURE IN RURAL AREAS
The agriculture production systems in the rural communities of South Africa are complex
and have developed under the influence of the climate, labour, economics, traditional
beliefs, and the communal land tenure system. Due to maize being the staple diet, the crop
planted is mainly maize while other crops, such as pumpkin and cowpeas, are intercropped
or rotated. In some areas crop rotation is practised although maize is mostly planted every
year. The seed used the following season is generally the previous year’s seed held over,
which is graded by the selection of the larger kernels of the harvested maize. When seed
is bought it is generally of the older hybrid variety, such as Pioneer 4141 in Mlondozi
(Jansen van Rensburg 2002). In areas such as Hluhluwe KwaZulu Natal, Round Up
Ready® seed is planted by the farmers, who are normally older and who control weeds by
spraying with Round Up®, to avoid the manual weed control.
The method of planting varies from community to community. Soils are prepared by animalor tractor drawn ploughs, where the ploughing depth is shallow. This is due to the limitation
of the plough design as well as to the animals being weak after the winter, so that the plough
cannot be pressed deep into the soil when ploughing. The tractors used are generally
poorly maintained, which includes having well-worn rear tyres, resulting in poor traction.
After the soils are ploughed the soil is smoothed over by the use of hand hoes. Planting is
done after furrows are drawn, using hoes (Jansen van Rensburg 2002). If fertilizer is used,
it is spread in the furrow and covered with a thin layer of soil, and the seed is dropped in
the furrow and covered with soil. In some districts, such as Giyani, the seed is broadcast
over the soil then ploughed in.
18
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
The fertilization of the crop on the traditionally planted fields is mostly with ‘kraal’ manure or
a mixture of ‘kraal’ manure and chemical fertilizer. The nutrient value of the manure varies
greatly due to the quality of the grazing, the bedding, whether the liquid excrement has been
retained and the amount of leaching that has taken place in the ‘kraal’. In a study conducted
in the Eastern Cape the nutrient content of dairy manure varied considerably as indicated
in Table 2.1. From the table the nitrogen content varied between 1,4 kg to 9 kg per ton of
manure. (Jezile, 2004)
For the rural farmer to supply the recommended 200 kg of 2:3:2
(22), two ton of manure per hectare would need to be applied. On the traditional plots, the
applied quantity of manure was far below the recommended amount.
Table 2.1 Nutrient composition of dairy manure
Manure type
Total Nitrogen
kg ton-1
Phosphorus
kg ton-1
Potassium
kg ton-1
Manure scraped from
‘kraal’
4,5
2,7
4,1
Range of nutrients
1,4 to 9
0,3 to 5,8
0,9 to 9
Weed control is made difficult by the fact that there is intercropping. The weeding is done
in December and in some areas a follow up weeding in January. Harvesting is made difficult
as the result of certain weeds germinating late in the season like the Black Jack (Bidens
pilosa). Women would pull large plastic bags over themselves to protect themselves from
the Black Jack seeds when harvesting.
Harvesting is done manually, by picking the cobs by hand, by pulling the sheath leaves
apart and twisting the cob off. The cobs are stored in a slatted wooden structure that is
raised off the ground. Due to the slats being wide apart the cobs are able to dry off to the
desirable 12% moisture content to prevent harmful fungal growth such as Diplodia ear rot
(Diplodia maydis) or Fusarium ear rot (Fusarium moniliforme). Older women in the family
19
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
unit do the threshing. The method followed is by rubbing the cobs against each other or
against a stone. The threshed seed is stored in bags with ash sprinkled over the seed, to
prevent possible insect damage (Jansen van Rensburg et al, 2001, 2002, Jansen van
Rensburg 2002).
D’Haese & Kirsten (2006) describe land tenure as the relationship of one person to another
with respect to land resources. These motivate production efforts by the land occupier. For
example farmers would not plant a crop if it were to be taken by a third party. Tenure
determines who can do what with the property in question and under what circumstances
they can do it. Tenure is often misunderstood as defining relationships between people and
property, whereas in fact, tenure defines social relationships between people.
Most African farmers hold their land under indigenous usufruct land tenure systems
irrespective of the formal legal position under national law. Evidence has shown that land
titling and registration has not yielded positive benefits. On the other hand, there is growing
evidence that indigenous tenure systems are dynamic and evolve with changing social,
economic and political circumstances. Rukuni (1998) argues that traditional or customary
tenure systems offer as much security as any other system provided that communities have
legal ownership and authority over their land and natural resources.
Under communal land tenure individual farmers have no incentive for implementing
sustainable land use systems.
The farmers are given land to farm, but without full
ownership. There is the proviso that if the land is not worked there is a chance of losing the
land. As a result farmers are reluctant to invest in the land by liming or fertilizing. Arable
land is usually allocated to individual families but there is no security of acquiring loans for
inputs (Jansen van Rensburg et al, 2001, Laker, 2012). The net result is the underutilization
of high potential land, with resultant unbearable pressure on marginal land and often a
20
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
serious degree of degradation. The farming enterprise is furthermore made difficult by the
distance to large centres.
Extension officers supply a limited amount of support to farmers due to their poor technical
training. In performing their day to day tasks, support is limited by aspects such as
availability of vehicles, as well as limited kilometres to visit farmers. In KwaZulu Natal
(personal communication, 2010) some extension officers would take a taxi, at their own
expense, to attend meetings.
LandCare was initially developed in Australia around 1997, to have a community based
approach for the raising of awareness, influencing farming and land management practices
and delivering positive environmental outcomes (Klucas, 1999). LandCare was adapted for
South Africa as reflected in the goal of the National LandCare Program to develop and
implement integrated approaches to natural resource management. The approaches were
efficient, sustainable, equitable and consistent with the principles of ecologically sustainable
development. LandCare, in the ARC, was initiated with a project in Mlondozi, a community
on the western border of Swaziland, followed by the Bergville LandCare project. The long
term and short term goals are outlined in the Department of Agriculture LandCare Policy
document (Molope, 1999).
LandCare success is attributed to: “… the universal recognition that the best way to tackle
the on-ground problems is to have the community deciding what their local priorities are”
(Klucas, 1999). This approach gave the community the opportunity to take ownership of
their project. Community development is generally a slow process, including the need for
development of monitoring and evaluation: The three criteria targeted for monitoring and
evaluation were social acceptability, bio-technical feasibility and economic viability. At the
end of the project an Impact Assessment Survey was conducted which showed the project
21
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
had a significant impact on the community (Jansen van Rensburg, Engelbrecht & Smith,
2003).
There are a number of doomsday messages pertaining to soil erosion in South Africa which
are regarded as “often misguided” by Garland, Hoffman & Todd (2000).
Scientific
assessments indicate the situation may not be as severe as popular literature suggests.
Popularly quoted figures of sediment yield including by Midgley (1952), Schwartz & Pullen
(1966) and Rooseboom (1976) are of 363 million tonnes (3 t ha-1yr--1); 233 million tonnes
(1,9 t ha-1yr-1) and 100 to 150 million tonnes (0,82 – 1,22 t ha-1yr--1) respectively. These
figures are based on sediment yields of main rivers. There are other published values which
are less reliable, due to the methodology followed to determine the figures, which was not
often clearly explained. The mean Africa value is 7,15 t ha-1yr-1 (Garland, Hoffman & Todd,
2000).
South Africa is largely semi-arid, with most of the land affected by agriculture which means
the figures are regarded as not being unreasonable. South Africa’s erosion figure is lower
compared to all continents except Europe. It is to be noted that these are sediment yields
which are not directly comparable to soil loss, although there will be a relationship. Soil loss
is the quantity of soil lost at a point source or catchment, while the sediment is the quantity
of soil transported by water which can be as much as 5 times lower than soil loss (Scott &
Van Wyk, 1992).
2.2 CONSERVATION AGRICULTURE
There is an increasing number of farmers adopting Conservation Agriculture as a farming
system, which includes No-Till (Horowitz, Ebel & Ueda, 2010). The beneficial effects of
conservation agriculture include reduced erosion, and run off, increased infiltration (Aon,
22
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
Sarena, Burgos, & Cortassa, 2001; Stubbs, Kennedy & Schillinger, 2004; No-Till,
Advantages and benefits in crop production, 2005; Chivenge, Murwira, Giller, Mapfumo &
Six, 2007; Diallo, Boli & Roose, 2008, So, Grabski & Desborough, 2009).
Further
advantages include the retention of nutrients due to the charges on humus (No-till club,
2005). This would be due to the increased organic carbon and mycorrhiza concentration
(Jehne, 1981). In West Africa, for instance, the organic carbon content of the soil is used
as an indicator of soil health (Bationo & Buerkert, 2001). The ‘Cornell soil health training
manual’ has organic content as one of its 39 indicators, for the rapid assessment (Gugino
et al, 2009).
The aim of implementing Conservation Agriculture is to improve crop yields while reducing
production costs, maintaining or improving soil fertility and conserving water, in other words
implementing an improved and sustainable farming system. It is a system that can be
adapted to a variety of farming situations. In its original form it is based on three principles
(FAO, 2001; FAO, 2007):
•
Disturb the soil as little as possible
•
Keep the soil covered as much as possible through dead or living mulch
•
Crop rotation
Furthermore other farming processes, such as weed and pest control as well as nutrient
management, are addressed in an integrated manner. Weed control is through the use of
herbicides, along with cover crops. The cover crops brought into the system are chosen to
rectify problems such as soil compaction or increasing soil organic carbon levels, in the
system. Farmers in Dakota have found that by bringing in a winter cover they have reduced
the weed pressure as well as the amount of fertilizer required to zero, in some cases (Video:
The next step, 2013). Other problems controlled by cover crops include:
•
Japanese radish for compacted soil
23
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
•
Cabbage and Sunhemp to control nematodes
•
Oats as a cover crop to reduce weed pressure
•
Black oats to improve phosphorus levels
The implementation of Conservation Agriculture improves soil health more holistically,
including physical, chemical and biological properties. Farms such as the ZZ2 estates are
using the “Cornell soil health assessment training manual” (Gugino, Dowu, Schindelbeck,
van Els, Wolf, Moebius-Clune, Thies. & Abawi, 2009) to test their soils as to their farm’s soil
health. Over the years the concepts and understanding of the soil’s physical and chemical
properties have been well accepted. It is not until recently that the improvement of soil
biology has become the focus. The definition adopted by the Cornell team is: “Soil health
is the concept that deals with the integration and optimization of the physical, chemical and
biological properties of soil for improved productivity and environmental quality”.
The
characteristics of a healthy soil include good soil tilth, sufficient depth, sufficient but not
excess nutrients, large population of beneficial organisms, free of chemicals and harmful
toxins and a small population of plant pathogens and pests.
Their approach has been to address soil degradation matters that result in degraded soils
which in turn result in low crop yields. The issues include soil compaction, surface crusting,
and low organic matter. There are thirty nine indicators listed in the rapid assessment for
soil health. The indicators are divided into three categories, namely physical, biological and
chemical.
Traditionally soils have been ploughed to control weeds and improve the soil structure for
the seed bed. In reality, in the long term, the soil structure has been gradually broken down,
along with a reduction in organic carbon. The practice of CA keeps the soil covered by the
crop residue or cover crops to reduce erosion by water and wind.
Crop rotation is
24
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
incorporated into the system to improve soil fertility as well as control pests and diseases
(Stubbs, Kennedy & Schillinger, 2004).
Cultivated soils’ susceptibility to erosion is demonstrated by studies that indicate that soil
erosion is 3 to 15 times higher on recently tilled soils compared to untilled soils (Box &
Bruce, 1996). An aim when implementing CA as a farming system, is to reduce soil erosion
due in part to the increased organic content (Kundu, Bhattacharyya, Prakash, Gosh &
Gupta, 2007) which would be beneficial especially in South Africa, which is a country with
a generally low soil organic content (Barnard, van der Merwe, De Villiers, van der Merwe,
& Mulibana, 2000). Observed benefits of CA by Jan Dube, a farmer in Mlondozi was the
reduced frequency of irrigation required, and reduced insect damage in his vegetable
garden (personal communication, 2003).
A minimum of 30 % soil surface covered with crop residues is required in conservation
tillage. As the cover approaches 100% soil erosion approaches zero. With a 50% soil cover
the erosion is reduced to by about 83 %, and 10% cover the erosion is about 30%
(Nyakatawa, Reddy & Lemunyon 2001).
In a cotton study conducted at the Alabama Agricultural Experiment Station, with kaolinitic
soils, three cultivation systems were implemented i.e. conventional, mulch-till and no-till;
two cropping systems i.e. cotton in summer and fallow in winter and summer cotton followed
by winter rye; and three nitrogen levels i.e. 0, 100 and 200 kg N ha-1; and two nitrogen
sources: ammonium nitrate and chicken litter. Measured differences for the treatments
included:
•
The no-till and mulch-till system produced on average 10 cm taller cotton than the
conventional tillage system
•
Higher cotton lint yield under no-till compared to conventional till
25
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
•
The soil moisture measurements for the top 7 cm were higher for the no-till plots
compared to the conventional tilled plots with and without winter rye cover crop.
The increased soil water resulted in earlier seed emergence, seedling vigour and
plant growth.
•
Soil erosion for no-till was about one third that of conventionally cultivated soils
(Nyakatawa, Reddy & Lemunyon, 2001).
The reduced erosion in CA farming systems can be attributed to the increased aggregate
stability due to increased mycorrhiza hyphae concentration (Goddard et all, 2008, Kladivio,
2001). Arbuscular mycorrhiza fungi hyphae lengths range from 3 to 30 m g-1 of soil, but
have been measured to be 68 to 101 m g-1 of soil in undisturbed grassland (Jones, Nguyen
& Findley, 2009). Soil aggregate stability is influenced by Basidiomycetes mycelia that
excrete polysaccharides.
Stable clay-humus complexes further good structure and
increase water reserves. (Husson, 2003). Soil structure is further improved by the root
exudates on soil particles which also improve soil structure by increasing aggregate stability
(Beare, Hendrix & Coleman, 1994).
There is a reduction in soil bulk density due to the implementation of No-till along with
controlled traffic. A study conducted in Australia on a soil cultivated for a period of 100 years
of conventional tillage, had a 22 month period of controlled traffic.
There was an
improvement of bulk density from 1,4 g cm-3 to 1,25 g cm-3. The available water capacity
improved from 10,2 mm per 100 mm soil depth to 15,4 mm per 100 mm depth of soil
(McHugh, Tullberg & Freebairn, 2009).
2.3 FACTORS AFFECTING SOIL STABILITY AND INFILTRATION
When water is applied to soil, whether in the form of precipitation or irrigation, some water
will flow into the soil. The remaining water will not penetrate the soil, resulting in pooling on
the surface or runoff occurring. Infiltration rate is defined as the volume flux of water flowing
26
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
into the profile (Shainberg & Levy, 1996). The decrease in infiltration can be due to the
matric suction gradient which occurs as the infiltration proceeds. Secondly the reduction
could be due to the deterioration of the soil structure caused by the impact energy of water
drops which leads to partial sealing of the profile.
Soil erosion is a complex process of interactions of the detachment and transport of soil
particles by raindrop impact and overland flow and deposition.
Factors affecting
detachment include raindrop size and shape, force and impact stress (Bradford & Huang,
1996).
2.3.1
Crust formation
Crust formation is due to the impact of rain drops on the soil surface, where the soil particles
detach from the aggregates, to form a dense layer in the top part of the soil surface. This
in turn results in reduced infiltration and increased run-off (Morin, Benyamini & Michaeli,
1981). In brief, crust formation is due to three mechanisms:
•
Physical breaking down of soil aggregates by the force of the rain drops and
•
Physiochemical dispersion and movement of clay particles.
•
A third mechanism proposed is the reduction in water movement into the soil and
the swelling of clays which could block the conducting pores, as suggested by Quirk
and Schofield (cited by Levy, 1988).
The particles clog the pores in the soil surface (Aggasi, Shainberg & Morin, 1981). The
dispersion of clay particles could be affected by the exchangeable cation concentration and
composition in the soil.
Thus the stability and reversibility of wet crusts formed is
dependable on whether the crust (or seal) is fully or partially formed. Rainfall of high
intensity and of long duration would result in a fully formed crust. A high intensity short
27
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
duration or low intensity long duration on the other hand would result in a partially formed
crust (Aggasi, Shainberg & Morin, 1988).
2.3.2
Parent material
Parent material is arguably the most dominant of factors that determines the respective
soils’ characteristics, including erodability. The Department of Water Affairs (1986) put out
a map of sediment production for South Africa, which demonstrated the close relationship
between the geology and sediment yields. Studies in South Africa have included the
collecting of soils and testing the erosivity in a laboratory rainfall simulator, showing that
certain soil types are more susceptible to erosion than others (Stern, 1990; Rapp, 1998).
Parent material is the fundamental building block of soils. Soils formed from basic parent
materials, such as dolerites, are more stable than those from acidic parent material, such
as granites, as indicated by the erosivity values. Smith (1990) studied soils developed on
basic and igneous rocks with respect to stability and degree of weathering. The study
focused on red soils which are inclined to be more stable due to higher free iron, aluminium
and organic carbon content (Smith, 1990). Conclusions made by Smith were that the most
stabile were those soils derived from basic rocks such as dolerite, and secondly soils high
in kaolinites.
As the result of excessive erosion taking place in the Transkei, D’Huyvetter (1985) focused
a study on establishing slope criteria for the prediction of erosion for different soil types.
Soils derived from dolerite showed higher stability against erosion compared to mudstone
derived soils. It was concluded that the role of parent material was vital in predicting erosion
potential. In many areas of South Africa the “planners” implemented a blanket rule of
ploughing all soils with a slope of 12% or less, without taking into account the soil type. The
maximum threshold rule for slope percentages for arable land for major soils was compiled
in the study by D’Huyvetter (1985). Shortlands was the only soil that had a maximum
28
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
threshold value of 13%, with Hutton at 10,5%. On the opposite end of the scale the duplex
soils such as Escourt and Sterkspruit had a maximum slope of 4,5%; and Swartland 5,5%.
This is further elaborated on in section 2.3.8.
2.3.3
Degree of weathering
D’Huyvetter (1985) and Weaver and Van Breda (1991) both found that the degree of soil
erosion was inversely correlated with the average yearly rainfall. Laker (2004) pointed out
that increased rainfall leads to more advanced soil formation, resulting in more stable soils,
resulting in less erosion. Van der Merwe, De Villiers, Barnard, Beukes, Laker, & Berry
(2000) found the erodability of melanic soils also decreased with increasing degree of
weathering.
Weathering is a combination of physical and chemical alteration of material, which must
occur for soil formation to take place. Physical weathering would take place, such as cracks
forming in rock due to uneven thermal expansion or contraction. Water is an essential
element for all chemical and physical weathering processes. Water is furthermore involved
in profile formation through percolation, evaporation, erosion (run off) and stagnation
(Schroeder, 1984). Smith (1990) found that the degree of weathering is related to the rainfall
of the district. The end products of acidic parent material of high rainfall areas have higher
percentages of clay and predominantly kaolinites, quartz and mica in the clay fraction of the
soil (Smith, 1990).
2.3.4
Free iron and aluminium oxides
Soil from the most highly weathered basic parent material, containing the highest amounts
of organic carbon, clay, free iron and aluminium, was also the soil that maintained the
highest Final infiltration rate and cumulative infiltration values under cultivation (Levy, 1988;
Smith, 1990).
29
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
2.3.5
Clay mineralogy
In Gauteng and Mpumalanga kaolinites are the dominant clay mineral, resulting in soils that
are known to form a seal and produce much runoff and erosion when exposed to rainstorms.
The hydraulic conductivity (HC) of a soil is correlated to the soil texture, mainly the clay
content. The higher the clay content the lower the HC. This is further affected by the
sodicity, or ESP (Exchangeable sodium percentage) content of a soil (Levy, 1988). The
most unstable soils are those with a high percentage of clays with a 2:1 layer of silicates,
especially montmorillonite, while the most stable are those high in kaolinite (Levy, 1988).
Hutton soils with predominantly kaolinitic clays are relatively stable. The stability is affected
by sodium and interstratified minerals such as smectites. Soils with smectites and illites are
more dispersive than soils with dominantly kaolinites (Goldberg & Glaubig, 1987; Levy
1988).
2.3.6
Sodium and magnesium
The degree of stability of a soil against disaggregation or dispersion determines its
resistance to erosion. There are two main factors that bring about structural stabilization
namely organic matter and iron oxides (Sumner, 1957 and Smith, 1990). South African
soils’ organic matter is generally low and even lower when cultivated. A number of soils
lose their structure when cultivated due to breakdown of organic matter. Soils of a dolerite
parent material retain their structure due the cementing effects of iron oxides. Smith (1990)
found in his study that soils derived from acidic igneous rocks were less stable than the
soils derived from the basic igneous rocks.
D’Huyvetter (1985) found a significant correlation between the ESP of the soil and the
degree of erosion. He found the highest ESP values for the Valsriver, Vilafontes and
Westleigh forms, all of which have a high erodability. The more stable soils such as the
30
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
Hutton, Shortlands and Oakleaf had lower ESP values. A high ESP resulted in dispersion
taking place causing dense crust formation, which reduces infiltration of water into the soil.
2.3.7
Organic material
It well known that organic material has various benefits which include improved soil
structure (Abiven, Menasseri & Chenu, 2009). In West Africa where soils generally have a
low organic carbon content, soil fertility is “defined” by the amount of carbon in the soils
(Bationo & Buerkert, 2001). Organic material binds micronutrients and metal ions that
otherwise might be leached out of the surface soil (Stevenson, 1994, Laird, 2001).
In a study compared conventional tillage to various direct drilling systems, Diallo, Boli &
Roose (2008), constructed two sets of run-off plots (57 in Cameroon and 17 in Mali) on
rather fragile sandy Alfisols under Sudanese savannah areas. After 3 to 4 years with 900 to
1500 mm of annual rainfall, the litter/legume/weeds/ cover had a reduced runoff of 20% and
33% reduced erosion compared to the conventional tillage system. The soil organic carbon
increased from 0,69% to 0,87% in a 3 year period. There was also found to be a reduction
in aluminium toxicity in the soil (Tang, Zang, Schroder, Payton & Zhou, 2007).
The practice of no-till is reported to sequester carbon. In a study the amount of carbon
sequestered was found to be 0,81 metric ton per ha for minimum tillage, compared to 1.58
metric ton per ha for No-Till (Horowitz, Ebel & Ueda, 2010). When fields are cultivated the
carbon is decomposed more rapidly. Loss of organic matter due to cultivation had an
adverse effect on infiltration rate values as well as the total infiltration into the soils. This
was further aggravated if a fallow period is included in the rotation (Tisdall & Oades, 1982,
Smith, 1990)
In a study conducted in Northern Mississippi soil erosion was found to be 47% and 50%
31
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
lower for plots with 50 and 100% plant populations respectively compared to the control
with 0% soil cover (Bot & Benits, 2005; Wilson, McGregor & Boykin, 2008).
There are a number of components that aid in aggregate stability including organic gels that
develop due to the transformation of organic matter (D’Huyvetter, 1985). The particles
would also be further stabilized by fungal hyphae coating the soil particles and the
extracellular polysaccharides produced by fungal hyphae (Tisdall, Nelson, Wilkinson, Smith
& McKenzie, 2012). The aggregate formation process continues by organic anions forming
complexes with metal ions which favours the aggregation of clays (Jastrow & Miller, 1998;
Laird, 2001).
Fungi play a major role in the stabilisation of soil, where it is estimated that up to 70% of the
carbon in the soil occurs in fungi. A by-product from the fungi is glomalin which may
polymerize and form hydrophobicity due to drying and exposure to air, thus protecting the
hyphae from desiccation.
Wessels (1996) proposed that the glue-like property and
hydrophobicity aids the stabilising of aggregates exposed to rapid wetting and drying.
2.3.8
Slope
In the Ciskei during the 1970’s there was alarmingly high erosion taking place in the so
called “betterment schemes”. Hensley & Laker (1975) identified the cause to be standard
slope criteria of 12% for arable soils by “planners” without taking into account the
erodabilities of the respective soils. Soil erosion is of importance because, the higher the
degree of erosion, the lower is the yield. In a study conducted in Michigan, there was an
average of 21% yield reduction on severely eroded soils compared to slightly eroded soils,
over a five year study period (Mokma & Sietz, 1992). D’Huyvetter (1985) focused a study
on establishing threshold slope criteria for the dominant arable soils in three pedosystems
of the former Ciskei. For each soil form a regression analysis was carried out with the
32
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
degree of erosion as a dependent variable and slope gradient as independent variable.
Theoretical threshold slope percentages for arable fields were determined. Significant
differences in erodability were found for different kinds of soils. The result was that each
soil form had different threshold criteria (Table 2.2).
Table 2.2 Recommended slope percentages for the arable soils of Mavuso
(D’Huyvetter 1985).
SOIL FORM
THRESHOLD PERCENTAGE %
Glenrosa
6,0%
Hutton
10,6%
Oakleaf
6,1%
Shortlands
13,0%
Swartland
5,5%
Escourt, Sterkspruit, Valsrivier, Vilafontes
4,5%
The parent material played a significant role in the stability of the soils. Hutton and
Shortlands which are derived largely from dolorite were the most stable, compared to the
majority of the other soils, mainly derived from Beaufort mudstone.
Stern (1990) conducted a study on unstable soils, which tended to seal and are highly
erodible, in a laboratory rainfall simulator where the slope of the soil could be accurately
set. The run-off from the control plot soils, set at a steeper slope (30o), was lower, with a
higher infiltration rate compared to soil on a gentle (5o) slope. This is explained by the crust
being eroded from the steeper slope, thus allowing water to infiltrate.
2.4
METHODS OF SOIL EROSION RESEARCH
Over time different methodologies have been developed to study soil erosion in the field
and relevant processes in the laboratory. Basic research was normally conducted on runoff
plots (1950-1960’s). From the 1980’s more diverse forms of apparatus were developed and
33
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
used. Researchers have had a need to study and quantify the effects of rainfall on the soil
by measuring soil erosion, run-off and infiltration. This study has been extended to the
effects of different farming systems on the soils erosivity. Studies have made use of runoff plots laid out in field, of varying sizes (Kongo & Jewitt, 2006; Jin et al, 2008). The water
source for field studies has been in the form of rain for in-field run-off plots, whereas distilled
water has been used for rainfall simulators. The rainfall simulators vary in size. (Swanson,
1965; Roth, Meyer & Frede, 1985; Miller, 1987)
Historically in-field run-off plots were the preferred method used to determine erosion from
soils. Plot size has traditionally been large (>10m2). Large plots are more suited to
hydrological studies (Kongo & Jewitt 2006), and include erosion and pesticide movement in
runoff from irrigated cotton (Silburn, Hargreaves, Budd & Granville, 1996; Silburn, Waters,
Connolly, Simpson & Kennedy 1998), and nutrient movement from dairy effluent applied to
pastures in southern and northern Queensland (Loch et al. 2001).
Studies have made use of run-off plots laid out in field, of varying sizes (Munn & Huntington,
1976, Roth, Meyer & Frede 1985, Miller, 1987, Kongo & Jewitt, 2006; Jin et al, 2008). There
is a recent trend for researchers to study runoff on smaller plots (<1m2) (Roth, Meyer &
Frede 1985; Claassens & van der Watt 1993; Loch, Connelly & Littleboy 2000, Seeger,
2007). The smaller plots are suited to gathering of data for modelling objectives (RUSLE,
1993). This is in line with the runoff studies using laboratory rainfall simulators.
The common method used by researchers for soil sampling for rainfall simulator studies
was to use a spade to dig out the topsoil which was followed by drying and sieving of the
soils, from literature cited (Levy, 1988, Smith, 1990, Stern, 1990). The dried and sieved
soils were placed in perforated trays for the simulated rainfall event. No reference was
found where undisturbed samples were taken for simulated rainfall events using a
laboratory rainfall simulator. Undisturbed soils tested through the use of portable rainfall
34
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
simulators are expensive due to the method, and are also time consuming. The motivation
for the taking of undisturbed soil samples was to obtain samples that were more
representative of the in-field situation. In most previous laboratory rainfall simulator studies
a single large sample was taken from a point source, representing each soil type.
This study complements the erosion studies conducted previously by having multiple
samples of the same parent material that were taken with minimal disturbance. The listed
studies consisted of single samples of each soil type. The studies conducted in South Africa
on soil erosion include:
1)
Sumner, M.E. (1957) The physical and chemical properties of tall grassveld soils of
Natal in relation to their erodability
2)
De Huyvetter, J.H.H. (1985) Determination of threshold slope percentages for the
identification and delineation of arable land in Ciskei
3)
Levy, G.J. (1988) The effects of clay mineralogy and exchangeable cations on some
of the hydraulic properties of soils
4)
Smith, H.J.C. (1990) MSc dissertation – The crusting of red soil as affected by
parent material
5)
Stern, R. (1990)
Effects of soil properties and chemical ameliorants on seal
formation, runoff and erosion
6)
Bloem, A.A. (1992) Criteria for adaption of the design and management of overhead
irrigation systems to the infiltrability of soil
7)
Rapp, I. (1998) The effects of soil properties and experimental conditions on the rill
erodabilities of selected soils.
35
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
CHAPTER 3
MATERIALS AND METHODS
3.1
SITE SELECTION
The project ran from 1999-2003. Eighteen farmers were randomly chosen (as unbiased as
possible) by the community in 1999 to be trained in the No Till farming system (Figure 3.1
and Table 3.1). The Conventional Tillage farming system was applied next to the NT
farming system, for comparative and educational purposes.
The selection of the farms was influenced by the availability of arable soils and forestry
plantations, particularly in the Southern half of the study area. The criteria for the selection
of farms included: the soils needed to be 600 mm plus in depth; the fields needed to be
fenced and easily accessible to the researchers and fellow farmers. At the start of the
project, each farm was measured and marked out with a quarter hectare each for NT and
CT adjacent to each other.
36
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
Figure 3.1 Mlondozi study area showing the distribution of the study sites.
37
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
Table 3.1 Names of farmers and co-ordinates for the 18 study sites
Numbering
of study
sites
Farmer name
Ward
Latitude (S)
Longitude (E)
1
Caiphus Dludlu
Belvedere
26o
12’ 7.5”
30o 49’ 59.1”
2
Jan Dube
Oshoek
26o 13’ 37.8”
30o 57’ 45.7”
3
Victoria Sheba
Oshoek
26o 12’ 28.2”
30o 58’ 3.8”
26o
30o 58’ 14.1”
12’ 57.3”
4
Josia Nkosi
Oshoek
5
Enoch Mavimbela
Oshoek
26o 12’ 30”
30o 56’ 22.1”
6
Vegetable garden
Mayflower
26o 29’ 1.5”
30o 45’ 35.7”
7
Amos Habile
Izindonga
26o 21’ 37.6
30o 46’ 45.0”
8
Paulos Shongwe
Ndanga
26o
21’ 52.4”
30o 46’ 10,5”
9
Joyce Simelane
Hereford
26o 17’ 48.4”
30o 48’ 12.5”
10
Albert Jele
Hereford
26o 17’ 45.2”
30o 47’ 58.4”
11
D.J. Nkosi
Hereford
26o 18’ 12.6”
30o 47’ 48.8”
12
Joseph Maseko
Hereford
26o
18’ 44.4”
30o 47’ 57.4”
13
James Makonsa
Syde
26o 17’ 47,9”
30o 48’ 38,2”
14
Meshack Mkwanazi
Syde
26o 28’ 11.4”
30o 43’ 24.4”
15
Absalom Makhubu
Syde
26o 29’ 1.5”
30o 45’ 35.7
16
David Ndlanganlandla
Syde
26o 28’
46.5”
30o 45’ 27.4”
17
Mbuti Mkhonza
Syde
26o 28’ 1.1”
30o 43’ 31.6
18
Filemon Matunjwa
Dumbarton
31o 18’ 29”
28o 52’ 42”
Sites 1 and 18 were researcher managed trials. The remaining 16 sites were managed by
the respective farmers.
3.2
PRODUCTION SYSTEMS PRACTICED IN MLONDOZI
Two methods of production were followed. The two systems referred to – Conventional
Tillage (CT) production system and “No Till” production system (NT). They are described in
more detail below:
3.2.1
Conventional tillage (CT) production system
Tillage - The traditional method of tillage in the community was with animal drawn
mouldboard plough or tractor drawn mouldboard plough. Both methods resulted in a
shallow ploughing depth of 10 to 18 cm. The minority of farmers budgeted for or had money
38
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
to have the soil disced to break up the clods to prepare a seed bed, after ploughing. The
seed bed was mostly prepared by hand using a hand hoe.
Seed – Most of the farmers planted their traditional seed, or seed held over from the
previous season’s harvest. If a hybrid was planted, the most popular was Pioneer 4141, an
old hybrid. Some farmers planted by tractor drawn planter, although most planting was
done manually using a hand hoe. The process followed was the making of a furrow and
placing the fertilizer or manure in the furrow, then closing it with a thin layer of soil. The seed
was then placed in the furrow and closed with the remaining soil.
Weed control - The first weed control operation was the ploughing of the soil, which
occurred after the first rains, while the second weed control process was mid-December,
which was by hand hoe. This is a much less destructive technology than mechanical weed
control by tractor. Thereafter there were no further weed control operations, because the
maize plant was regarded to be mature enough to compete with the weeds.
Fertilizer - The fertilizer application by the traditional farmers depended on the farmers’
income. No lime was applied by any of the farmers under the conventional system. The
fertilizer application rates were zero for low income farmers who did not have access to
manure; dry powdered ‘kraal’ manure; or a mixture of ‘kraal’ manure and a chemical fertilizer
such as 2:3:2. If a chemical fertilizer was applied, it was applied at a rate up to 100 kg 2:3:2
(22) ha-1, although the recommended rate was 200 kg ha-1. The application of manure as a
fertilizer is good, as organic material is supplied, along with macro and micro elements. A
large quantity of this would need to be applied i.e. at a rate of 2000 kg ha-1, to supply the
equivalent amount of nutrients of 200 kg 2:3:2 (22) ha-1.
Pest control - The pest control practice followed to control Stalk borer, (Busseola fusca)
was to plant late in spring. By planting late in spring, it was the farmers’ objective and belief
39
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
that the stalk borer flights would be missed, and in turn reduction in larvae damaging the
crop. The late planting led to other problems, however, such as flowering of the plants late
in the season and the cobs not reaching physiological maturity before winter.
Fodder - In winter, animals were allowed to graze on the crop residue, reducing the mulch
covering the soil surface.
3.2.2
“No Till” (NT) production system
The method of training of the farmers in the “No Till” agriculture farming system was by
farmer led demonstration trials. Sufficient inputs were given to the participating farmers for
a quarter hectare. The rationale was to take the risk away from the farmer for the new
production system, and produce sufficient food for the family and still have sufficient crop
to sell so as to buy more inputs to expand the area under the NT production system. The
inputs consisted of hybrid seed (5 kg per 0.25ha: for a plant density of 35 000 plants per
hectare); 50kg 2:3:2 (22) fertilizer, 25kg LAN, 1 litre Round Up®, 1 litre Bullet® and 40cm3
insecticide (Bulldock®), which was sufficient for two spray applications per 0.25 ha. A
knapsack sprayer was also provided to each participating farmer. The inputs were given
with the proviso that:
•
The farmer would follow the NT production system;
•
The researcher was able to do a yield estimate;
•
The farmer would plant an additional quarter hectare of maize following their
traditional method of maize production alongside the NT plot.
By planting maize following the two methods each farmer was able to compare the growth,
labour input for each system and the crops’ reaction to stresses such as the dry period
between precipitations.
40
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
Field preparation - The fields were firstly limed, at an application rate of 5 ton ha-1. The
fields were mechanically (tractor drawn) ripped to break up any compacted layer, and then
ploughed using a mouldboard plough to incorporate the lime. This was to correct the topsoil
acidity before starting with the NT farming system. The seedbed was prepared by a tractor
drawn disc, thereafter there was no vehicular traffic on the fields. The only soil disturbance
was planting, which was done by hand. This involved the making of a furrow 50 to 60 mm
deep for fertilizer and the planting of seed, using a hand hoe. The seed was placed in the
furrow at an intra-row spacing of 30 cm and row spacing of 90cm. The fertilizer was
measured out using a Coca Cola cool drink bottle lid (volume of 7 cm3) of fertilizer placed
in the furrow between every second seed. The seed and fertilizer was covered with soil.
Fertilizer – A general fertilizer recommendation for Mlondozi was 200 kg 2:3:2: (22) per
hectare. Limestone Ammonium Nitrate (LAN) was applied as a top dressing, six weeks
after planting, at a rate of 100 kg ha-1.
Seed - A Carnia hybrid, CG 2969, was planted at a plant density of 35 000 plants per
hectare.
Weed control – The programme followed was: Two weeks after the first summer rains,
Round Up® was sprayed at a rate of 4 l ha-1 to kill the young weeds, and retain the residue
as mulch on the soil surface. Planting took place seven days after spraying with Round
Up®. Within three days after planting, Bullet®, a residual herbicide, was sprayed onto the
soil at the rate of 4 l ha-1. The spraying of the herbicides was done using the supplied16 litre
Knap sack sprayers.
Pest control - Bulldock®, a pyrethroid based insecticide, was sprayed at a rate of 20 cm-3
per quarter hectare, every two weeks for a period of four weeks to control stalk borer larvae.
The spraying programme commenced when the crop was six weeks old.
41
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
3.3
CLASSIFICATION AND PROPERTIES OF THE SOILS OF THE STUDY SITES
At the start of the project a profile pit was dug in the two fields (per cultivation system NT
and CT) to classify the soils, at each participating farmer. This was to classify the soils and
sample with expectations that differences over time would develop.
The procedure to classify soils, as described in the book Soil classification: A taxonomic
system for South Africa (1991) is:
•
Demarcating the master horizons present in the profile
•
Identifying diagnostic horizons or materials
•
Establishing the soil form using the key
•
Identifying the family differentiae
•
Establishing the soil family
•
Determining the texture class of the A horizon and adding it to the name or code of
the soil family
3.4
THE LABORATORY RAINFALL SIMULATOR
The laboratory scale rainfall simulator at the ARC-ISCW used in this study, was identical to
that described by Morin, Goldberg and Seginer (1967) (Figure 3.2).
The simulator consists of two parts; namely:
•
The applicator;
•
The soil box carousel.
The applicator consists of a pump which supplies water at a given pressure to a rotating
nozzle and a rotating metal disc that can be altered for different apertures. The water drops
from the nozzle are able to spray out onto the boxes when the aperture is under the nozzle.
42
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
A high pressure rain was obtained when the water was at 60 kPa from a full cone spray
type no.1.5 H30 nozzle at a height of 2 m from the soil surface
A rain intensity of 45 mm h-1 was used, with a drop diameter of 1,9 mm diameter, with a
terminal velocity of the drop of 6,02 ms-1 and kinetic energy of the rain of 19,1 J mm-1 m-2
(Morin et al. 1967).
In previous work it was found that it was unnecessary to use soil layers thicker than 25 mm
since infiltration rates are unaffected even if thicker layers are used (Morin et al., 1967). A
piece of cloth was placed between the sand and the soil to ensure continuous flow of water
from the soil to the sand. The plastic trays containing the samples were placed in sealed
boxes containing sand. A hose was attached to a nipple on the underside. The sample
was wetted until saturation by tap water (EC = 1mS/m) forced through the sand, thus wetting
the sample by capillary action. When the sample was saturated the water supply was
switched off and the simulated storm event commenced, using distilled water. The angle
of the samples on the carousel was 5 degrees. This was used as all the previous rainfall
simulation studies conducted were at 5 degrees, thus the results could be compared to
previously conducted rainfall simulations.
The samples were placed in the rainfall simulator for two storm events. The first storm was
to settle the soil sample. The trays with the soils were removed after the first storm, dried
for seven days and then replaced in the carousel for a second storm event. Each storm
lasted for 50 rotations, with each rotation taking 2 minutes, thus each storm lasted for 100
minutes, giving an application of 75 mm per storm.
The amount of water that infiltrated the soil was recorded for every second rotation, along
with the runoff. The run-off was collected in volumetric cylinders, measured, and then
transferred to beakers, of which the mass had been pre-determined. A teaspoon tip of
43
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
magnesium chloride (MgCl 2) (about one gram) was added to the slurry and mixed in, to aid
flocculation of the suspension. The excess water was decanted, and then the slurries in
the beakers were dried for 48 hours in an oven at 80 oC. The standard temperature of 105
o
C, was not used due to the beakers being made of plastic (ARC-ISCW QMS procedure,
2007).
Figure 3.2 Schematic diagram of the main parts of the laboratory rainfall
simulator (Morin, Goldberg and Seginer, 1967)
3.5 SOIL SAMPLING FOR THE RAINFALL SIMULATOR
44
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
The objective of the pan method was to take undisturbed topsoil samples for a laboratory
rainfall study.
Secondly a faster turnaround time was sought under more controlled
conditions using the laboratory rainfall simulator.
The pan method of taking undisturbed soil samples developed in this study proved to be a
good alternative to the traditional methods of soil loss determination and run-off achieving
all of the above objectives. The traditional method of taking soil samples was digging out
the soil using a spade, then sieving the soils, which were tested in a rainfall simulator.
3.5.1
Sampling equipment
Mild steel sampling trays were constructed to take undisturbed topsoil samples. The trays
were made from 2 mm metal sheet. The metal was bent to shape to the dimensions of 350
x 500 x 50 mm. An opening was left on one long side (Figure 3.2).
Figure 3.2 Metal tray used to take samples in field
45
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
3.5.2
Sampling procedure
•
A shallow hole was dug, one and a half times the width of the tray and about 60 mm
deep. The face of the soil was squared so the soil would be flush with the inside of
the tray.
•
The tray was tapped horizontally into the soil using a 2 kilogram hammer, with
minimal disturbance, for a constant thickness of 50 mm.
•
After sampling the open edge was carefully cut with a spade for minimal disturbance
of the sample.
•
The metal boxes were wrapped in plastic cling wrap after the samples were taken.
•
On arriving back at the Institute from the field, the plastic wrap was immediately
removed to prevent possible decomposition of the mulch and to allow the sample to
air dry.
3.5.3
Transferring the soil samples to the plastic trays
Transferring of the soil samples from the metal trays in which the samples were collected
to the plastic trays used in the rainfall simulator was done as follows:
•
The soil samples were wetted with deionised water, using a 500 ml plastic squirt
bottle (with the squirt attachment removed so as to supply a low energy jet of water)
– for a friable sample.
•
A wooden laboratory tray was placed, upside down, over the sample while still in the
metal sampling tray. The sample was then flipped over so the top of the sample lay
on the wooden laboratory tray, with the sample lying upside down. The underside
was scraped to achieve a constant thickness of 50 mm and a flat bottom surface.
•
A muslin cloth was placed in the bottom of the plastic perforated rainfall simulator
tray to prevent soil seeping through the holes during the simulated storm event.
46
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
•
The perforated plastic tray was placed, upside down, over the sample that had been
worked on the laboratory tray. The sample was then flipped over into the perforated
plastic tray.
•
Along the edges of the sample, where there was not a good fitting of the sample in
the plastic tray, the soil that was scraped off, was pressed into the gaps (Figure 3.3).
•
A mask was made to cover the soil that had been disturbed along the edges, so as
not to influence the results. This was made using 50 x 50 mm angle iron, 3 mm thick.
A 2 mm diameter round rubber strip was glued on the edge of the angle iron closest
to the soil to prevent possible accumulated water running onto the soil surface
(Figure 3.4). The resultant area of the exposed soil surface was 240 x 470 mm = 0,
1128 m2.
Figure 3.3 Soil sample transferred to the plastic tray
47
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
Figure 3.4 Example, of a perforated plastic tray
3.6
•
YIELD ESTIMATION
Yield estimation was conducted yearly during the winter months of June or July
during the study period. Yield estimates were conducted in the NT and CT plots
as well as the researcher managed trial plots. Only accurate yield data was used
for data analysis due to the fact that some farmers harvested before time. The
inter-row spacing was 90cm and the intra row spacing was round about 31cm.
The procedure for yield estimation was as follows:
•
A row in the field where the yield estimation was to be conducted was randomly
chosen. The row had to be at least three rows in from the edge of the field, and
three meters from the end of the row, to avoid any boundary effect.
•
A five meter length of the row was measured and marked out.
•
The number of plants in the five meter row was counted for plant population.
•
The cobs were picked then placed into a large plastic bag which was previously
marked along with the details of the plant numbers.
48
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
•
This process was done in triplicate in each field.
•
At the laboratory the cobs were threshed and the moisture content determined;
weighed; and data recorded in the field book.
•
All the yields were converted to 12,5% moisture in the seed so as to have
comparable results.
•
Correction of seed water content to 12,5% moisture
= 1 −
•
%
,%
×
Formula for yield per hectare (kg per ha)
=
!"#$%×&"$!''()$%
×
*+,,-,!!)+$.%(,$/'!
3.7 DETERMINATION OF IN-FIELD MULCH COVER
The mulch cover has a direct relationship with the erodability of the soil. It has further
benefits such as reducing weed infestation and reducing evaporation from the soil surface
.Mulch cover percentage was estimated by using a series of photos with known percentage
cover of maize or soya beans. The sampling was done at the same time each year in
winter. The mulch cover was classified, using photos of a 25, 50, 75 and 90 percentage
cover. It is estimated that a person can be accurate to within 10 to 20% (Shelton and Jasa,
1998; Eck and Brown, 2004).
Photo-comparison method consists of:
49
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
•
A representative area randomly chosen in the field. Row ends or places where there
was an excess of weeds or other disturbances were avoided.
•
Making at least three observations, and then averaging the individual estimates to
obtain a percentage for the whole field.
•
3.8 DETERMINATION OF THE ROOT COUNT IN FIELD
When farmers adopted the NT farming system, it was expected that there would be a higher
root count per volume of soil for the NT fields compared to the CT fields. It was thus deemed
necessary to monitor the root count as a biological property to determine if changes took
place in the soils. The soils were monitored each year in winter for the top and sub soils.
A surrogate root count was used for comparative purposes. The method consisted of a
profile pit being dug in both the CT and NT fields. Root count was determined by placing a
20 x 20 cm piece of cardboard, with a previously cut out hole, of 10 x 10 cm, against the
side of the profile pit. This was done in the middle of the A and B horizons. The number of
roots observed in the window of the horizon were counted and recorded in the field book.
3.9
OTHER LABORATORY ANALYSES
3.9.1
Clay mineralogical composition was determined by X-ray diffraction (XRD), using a
CoKα source for x-rays and graphite monochromator. A semi quantitative evaluation of the
amount of the various clay minerals was obtained from the relative peak areas of the XRD
of the clays. This was done for completeness, but was not further elaborated on for the
purpose of this study.
50
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
3.9.2
The chemical analyses done on the soil samples followed the procedures described
in the handbook of standard soil testing methods for advisory purposes (The non-affiliated
soil analysis work committee, 1990). These procedures were as follows:
•
Extractable phosphorus: Bray 1, as described by Bray and Kurtz (1945)
•
Extractable cations: Ammonium acetate (1 mol dm-3, pH 7)
•
Organic Carbon: Walkley-Black method as described by Walkley Black (1934)
•
pH (H20); a 1: 2,5 soil - water ratio
•
Cation exchange capacity using lithium chloride.
3.10
STATISTICAL ANALYSIS
All the data used was gathered by the use of a random sampling system. Samples were
taken as described for each observation. The following statistical procedures were done:
3.10.1 Fertility
In winter each year topsoil samples were taken using an auger, to a depth of 30 cm. Three
samples were randomly taken per cultivation system (0,25 ha). The samples were mixed
together and a representative sample was taken for nutrient status analyses.
The plot layout (NT and CT) on the farms was as a split-plot design replicated eighteen
times (each farmer as a replicate). Analysis of variance (ANOVA) was used to test for
differences between the 2 treatment-systems (NT and CT) as main-plots and the 4 years,
1999-2002 as sub-plots, as well as for their interaction. The data was acceptably normally
distributed and were separated using Fishers' protected t-test least significant difference
(LSD) at the 5% level of significance (Snedecor & Cochran, 1980). The software
programme, SAS 9.2, was used to test the soil fertility elements (pH, P, Ca, Mg, K and
CEC).
51
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
3.10.2 Maize yield
The same statistical design, as well as analysis as described above, was applied to the
maize yield data. Method of yield determination is described under point 3.6.
3.10.3 Mulch cover and Root count
Tabulation of the mulch and root counts for each farming system (NT and CT) is presented.
3.10.4 Runoff, Infiltration and Soil loss
At the end of the project, four undisturbed soil samples were taken from each cultivation
treatment, using the metal trays described in 3.5. Soil sampling for the rainfall simulator
study was under-taken during July 2003, on moist soils.
3.10.4.1 Statistical Analysis part I
Statistical analyses for comparison of the two storms separately, were conducted on
sampled soils, to the soils for runoff, infiltration and soil loss on the data sets. The two
sample t - test was applied to the data to test for differences between the two systems
means and totals of the tests (sample size was = 72 at 18 sites with 4 replicates, (figure 4.5
and figure 4.6) The “students” t-test for two independent samples was applied to the data
to test for differences between the two systems means (sample size was = 72 at 18 sites
with 4 replicates). Significance was obtained at the 5% level of significance (P < 0.05;
Snedecor & Cochran, 1980). Data was analysed using the statistical programme GenStat®
(Payne et al., 2007).
52
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
3.10.4.2 Statistical analysis part II
The two storms were combined by analysing the data, as a split-plot analysis with the
treatment farming systems (NT and CT) as main plots and the storm and rotations as
sub-plots. This dataset was analysed with SAS 9.2.
53
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
CHAPTER 4
RESULTS AND DISCUSSION
4.1 Soil Classification and Mineralogy
The farmer’s soils were classified and listed, along with the horizon depths in Table 4.1. The
soils, although mostly having high carbon content in the topsoil, were not classified as humic
horizons due to the exchangeable cations being too high.
Table 4.1 Soil classification (Soil classification working group, 1991)
Sampling
point
1
Soil form
Soil series
Clovelly
Caledon
2
Clovelly
Lundini
3
Clovelly
Oatsdale
4
Hutton
Williamson 15
5
Hutton
Farningham
6
Clovelly
Oatsdale 12
7
Clovelly
Oatsdale 12
8
Kroonstad
Swellengift 12
9
Clovelly
Leeufontein 16
10
Clovelly
Oatsdale 16
11
Hutton
Hutton 16
12
Clovelly
Oatsdale 16
13
Avalon
Newcastle 25
14
Hutton
Hutton 16
15
Hutton
Hutton 16
16
Hutton
Hutton 16
17
Hutton
Hutton 16
18
Hutton
Hutton 16
Horizon A
(mm)
Horizon B
(mm)
200
650
220
800
220
800
260
600
300
520+
250
800
300
500+
200
700
260
600
280
800
250
800
250
800
250
550
300
1200
380
900
200
700
300
600
260
650
Clay mineralogical composition of the study site soils are listed in Table 4.2. The dominant
minerals are kaolinite and quartz. Common mineralogical end points after weathering are
54
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
goethite, gibbsite, hematite and coarse-grained quartz. In most samples there are a high
percentage of end-point minerals (0 to 34%), which is an indication of highly weathered
soils. (Jackson, Tyler, Willis, Boubeau and Pennington, cited by Mulibana, 2001).
Table 4.2 Clay mineralogical analysis of the soils from the 18 study sites (ARC –
ISCW).
Profile
no
Qz
Mi
Kt
Go
Gb
%
%
%
%
%
1
23
-
60
17
-
2
62
-
38
-
-
3
29
-
50
6
15
4
14
-
62
5
19
5
17
3
50
7
9
6
17
-
45
14
24
7
31
4
50
10
5
8
19
4
48
13
16
9
24
4
67
-
-
10
45
-
50
-
-
11
15
6
64
6
9
12
39
-
42
10
9
13
35
-
49
16
-
14
29
-
52
9
10
15
35
3
41
-
21
16
50
-
21
12
17
17
11
5
70
7
4
18
23
-
59
19
-
Qz – Quartz
Mi – Mica
Kt – Kaolinite
Go - Goethite
Gb – Gibbsite
4.2 Soil samples using the pan method.
The pan method was not compared with in-field run-off plots or an in-field portable rainfall
simulator. The constant similarity of the replicate infiltration rates indicates a method that
produces results that are constantly repeatable.
The simulated storm events, in the
55
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
laboratory rainfall simulator, deliver consistently precise simulated storm intensity and drop
size with purified water. The objectives achieved included:
•
Taking undisturbed soil samples
•
The samples were tested in the laboratory rainfall simulator for two simulated storm
events under controlled conditions.
•
The turnaround time was quick, making this a more economical method.
•
Within two hours after sampling the result could be obtained, using the rainfall
simulator.
4.3. RESULTS AND STATISTICAL ANALYSIS
4.3.1
Fertility
Due to soils being derived from granite and acidic in nature they have a low fertility, as
indicated by the base line soil samples of the 18 study sites taken in 1999 (Fig. 4.1- 4.3;
Table 4.3). Soil samples taken annually from the NT fields, to the end of the study period,
were compared to the base line samples taken at the start of the project. Although dolomitic
lime was applied at the start of the study period and there were no follow-up applications,
the Ca and Mg concentration and pH increased during the study period and were statistically
different at 5% level (between 1999 and 2002 - Table 4.3).
Although a 2:3:2 fertilizer
mixture was applied, potassium decreased over time, (significantly lower), but phosphorus
and carbon were not significantly lower between 1999 and 2002.
The fertilizer recommendation for the NT plots was aimed at improving soil fertility by
supplying the plant nutritional requirements and raising the base line nutrition levels in the
soil. A noticeable occurrence was an increasing trend in soil pH over the study period.
During the first season most of the extractable acid was neutralised and the soil pH (H2O)
showed a lag period of 2 to 3 years. The crops cultivated following the CT farming system
continued to have leaves that were yellow and purple due to nitrogen and phosphorus
deficiencies.
56
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
600.0
Average Ca and Mg concentration during the
study period
500.0
mg kg-1
400.0
300.0
Ca
200.0
Mg
100.0
0.0
1999
2000
2001
2002
Years
Figure 4.1 Average of the Ca and Mg, of the NT soils for the 18 study sites.
Average of K and P during the study period
mg kg-1
160.0
140.0
120.0
100.0
80.0
60.0
40.0
20.0
0.0
K
P
1999
2000
2001
2002
Years
Figure 4.2 Average of the K and P of the NT soils for the 18 study sites.
10.0
Average CEC status of the soils during study
period
cmol kg-1
8.0
6.0
CEC
4.0
2.0
0.0
1999
2000
2001
2002
Years
Figure 4.3 Average of the CEC of the soils in the NT fields of the 18 study
sites.
A combined ANOVA statistical analysis was done over the four seasons and farms, using
SAS 9.2 version (SAS Institute Inc., 1999) to test for differences (table 4.3). The data was
57
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
normally distributed with acceptable homogeneous variances over years. The means were
separated using Fishers' protected t-test least significant difference (LSD) at the 5 % level
of significance (Snedecor & Cochran, 1980).
Table 4.3 Combined ANOVA over seasons
Years
Ca
Mg
K
P
CEC
pH`
C
1999
2000
2001
2002
MSE
268.56 b
283.82 b
388.0 ab
494.68 a
36738.31
131.20 b
129.94 b
160.14ab
224.11 a
10336.78
137.19 a
61.76 b
79.89 b
76.35 b
1634.99
9.54 b
18.33 a
13.70ab
15.07ab
122.01
8.08 a
5.25b
6.64ab
.
5.950
5.22 c
5.49bc
6.03 a
5.92ab
0.391
1.975 a
1.897 a
1.849 a
1.887 a
0.043
F probability
0.0052
0.0375
<.0001
0.1407
0.0074
0.0008
0.802
LSD(5%)
134.72
70.5
27.92
7.706
1.77
0.432
0.149
1Means
1MSE
4.3.2
per column followed by a different letter were significantly different at the 5% level.
is the Mean Square Error
The Comparison of NT and CT farming systems - Yield
“No Till” (NT) is regarded as an improved farming system over the traditional farming system
(CT) and includes the use of correct fertilizer quantity and type, hybrid seed, herbicides and
insecticides, planting without ploughing and residue retention as mulch. In comparison the
traditional farming system used little to no fertilizer, seed held over or traditional seed and
planting late to avoid stalk borer damage.
Indicators used to gauge the two farming systems included yield. The yield increased from
0.51 ton per hectare at the start of the project to 4.6 ton per ha for the 2001 season and
2.76 ton ha-1 for the 2001/2 season. Following the implementation of the NT farming
system, the farmers observed that the crop on the NT managed fields was less prone to
wilting between rainfall events. That indicates a higher infiltration rate during rain events,
which would allow the soil to reach field capacity quicker, due to lower run-off. With a higher
infiltration rate, a larger reservoir of water was retained in the soil. In comparison the CT
58
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
farming system maize plants remained dwarfed with stunted root systems due to the acidic
soils, which resulted in the plants wilting relatively soon after rain events.
Table 4.4 Yields of the NT and CT field of the 18 farmers.
CT yields
NT yields
Farmer
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
Average
1999/2000
ton ha-1
1.9
4.074
2
3.829
1.304
2.253
1.674
2.463
3.288
2.556
5.089
4.319
3.656
1.7
2.86
2000/1
ton ha-1
3.7
7.22
4.26
5.33
6.41
4.26
4.88
2.13
3.39
3.11
4.8
4.64
5.2
4.52
4.4
4.94
4.6
4.58
2001/2
ton ha-1
6
2.88
2.88
2.4
2.88
2.88
3.36
3.36
3.36
2.9
3.29
1999/2000
ton ha-1
1.5
2.82
1.469
2.38
1.461
0.997
0.8
1.43
2000/1
ton ha-1
2
2.65
1.58
5.59
2.65
2
2.75
2001/2
ton ha-1
2.8
3.8
3.3
The yield estimation was conducted in July of each year. Some farmers harvested before
the researcher could do the yield estimation. There are 16 missing data points for the
2001/2 season because the farmers were gradually implementing the NT farming system
on the rest of their farms. The yield estimation was therefore, not representative for 2001/2
season as well the two previous seasons.
Due to the improved combination of technologies for NT, there was an improvement of yield
over the seasons and over CT. For 2001-2002 season there was a decrease in yield due
to a drier season.
59
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
At the start of the project the farmers agreed to continue cultivating maize on the traditional
plots following the traditional farming system. Over time they gradually started to implement
the NT technologies which is the main cause of the increased yield. NT and CT gave the
same yield results for 2001/2002 season, because of the dry season and higher fertilizer
input.
A split-plot (ANOVA) over years, treatments and farms was done using SAS 9.2 to test for
differences between years, treatments and years X treatment interactions (table 4.6-4.8).
The data was normally distributed with acceptable homogeneous variances over years. The
means were separated using Fishers unprotected t-test least significant difference (LSD) at
the 5 % level of significance (Snedecor & Cochran, 1980).
Table 4.5: ANOVA table for yields
Yield
Farm
D
F1
1
6
Treatment
SOURCE
MS2
P3
2.076
0.0742
1
16.128
0.0008
Error(1)
9
0.895
Year
1
10.314
.0009
Treatment X Year
2
2
5
5
5
4.646
0.0261
Error(2)
Corrected
1.097
Where DF1 = Degrees of freedom, MS2 = Means Square and p3= Significant level of F-Ratio.
A significant level less than 0.05 is considered as a significant effect and indicated in bold
Table 4.6 shows the mean yield for the two treatments, table 4.7 the mean yield over the
different seasons, for the study period and table 4.8 the mean yield for the two treatments
over different seasons.
60
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
Table 4.6 – Means of yield for the two treatments
Tmt
NT
CT
Yield1
3.68
2.30
N
41
15
a
b
LSD(p=0.05)= 0.65
Yield in ton ha-1
1
Table 4.7 – Mean yield over the different years for both farming systems.
Year
2000/01
2001/02
1999/2000
Yield
4.10
3.29
2.45
N
23
12
21
a
b
c
LSD1(p=0.05)= 0.74
1
LSD is the least significant difference.
Table 4.8 – Mean yield for different treatments over the seasons
Year
2000_01
2000_01
2001_02
2001_02
99_2000
99_2000
Tmt
NT
CT
NT
CT
NT
CT
Yield
4.58
2.75
3.29
3.30
2.86
1.63
N
17
6
10
2
14
7
a
bc
b
b
bc
c
LSD1(p=0.05)= 1.27
1
LSD is the least significant difference.
61
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
Average yield for NT and CT farming systems
5.00
Yield (ton ha-1)
4.50
LSD(p=0.05)=1.27
4.00
3.50
NT
CT
3.00
2.50
2.00
1.50
1999_2000
2000_2001
2001_2002
Seasons
Figure 4.4 Average of the NT and CT yields for the study period.
4.3.3
Mulch cover and Root count
Traditionally the farmers would allow the cattle to graze on the crop residue. By keeping
the cattle out of the NT fields and the farmers cutting down the residue after harvest of the
maize crop, there was a higher percentage of soil cover, going up from an average of 30%
cover, which was the base line estimation, to 55% soil cover (Table 4.9).
The FAO
recommendation for the implementation of NT is for the soil to have a minimum of 30% soil
cover. The retention of the mulch on the soil reduced evaporation from the soil surface,
reducing run-off resulting in improved infiltration. With reduced run-off there would be a
reduction in soil erosion taking place.
Table 4.9 The average mulch of the NT plots over the study period (Jansen van
Rensburg, 2002).
Baseline value at
Values in
Values in
Values in
initiation of project
2000
2001
2002
60
55
55
(1999)
% Mulch
30
62
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
The root count (Table 4.10) went up from 30 in 1999, to 41 in 2001 and dropped to 35 per
10 by 10 cm area in 2002, which was an overall 16% improvement in the top soil. For the
subsoil, the root count went up from 7 in 1999 to 16 in 2001, and dropped to 13 in 2002, still
an increase of 76% from the baseline in 1999. Again, for the NT treatments, there was an
improvement of root count due to the reduction of acidity due to liming. This in turn further
improved the nutrient and water uptake. Improved nutrient uptake resulted in an optimal
crop. An improved rooting system resulted in the crop not wilting as quickly as the crop
produced following the traditional farming system.
Table 4.10 Average root counts for the NT fields
Indicator
Baseline root
Root
Root
Root count
count at initiation
count in
count in
in 2002
of project (1999)
2000
2001
7
14
16
13
30
39
41
35
Roots per 100 cm2 –
subsoil
Roots per 100 cm2 –
topsoil
With there being an increase in yields there was an equivalent increase in biomass, which,
when left on the field, had improved percentage mulch and root counts. This resulted in the
observed reduction in erosion taking place. There is a correlation between the above-, and
below, ground root biomass. With an increase in below ground biomass there was an
increase in root counts in both the top and sub soil horizons for the NT crops. This was due
to the correction of the soil acidity and the breaking of the plough sole. Eighty percent of
the feeder roots occur in the top soil, which was supported by the retention of the mulch
due to the reduction of soil water evaporation. In turn the higher root count would lead to
an improved nutrient and water uptake and would potentially improve the quantity of organic
material, over time in the soil. There would be channels left after decomposition which
would further improve water infiltration.
63
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
4.3.4
Runoff, Infiltration and Soil loss for the NT and CT treatments
Two statistical analyses for above heading were performed.
4.3.4.1 Statistical Analysis part I
. The first part consists of: Two sample t - test was applied to the data to test for differences
between the two system means and totals of the tests (sample size was = 72 at 18 farms
with 4 replicates - table 4.11). Significance was obtained at the 5% level of significance (p
< 0.05; Snedecor & Cochran, 1980). Data was analysed using the statistical programme
GenStat® (Payne et al., 2007).
Table 4.11 Mean soil loss, run-off and total infiltration (n=72)
Soil loss (g)
System
NT
CT
Probability
Run off (cm3)
Total infiltration
(cm3)
Storm 1
Storm 2
Storm 1
Storm 2
Storm 1
8.46a
6.45b
0.043
7.39a
7.43a
0.965
1984a
1786a
0.127
1644a
1817a
0.22
2865a
3176b
0.015
Storm 2
3364a
3207a
0.291
Means within each column with the same letter or letters do not differ significantly at the
5% significant level.
Mean soil loss
9
Soil loss (g)
8
7
6
5
NT
CT
4
3
2
1
0
Storm 1
Storm 2
Figure 4.5 Graph of the mean soil loss
Fig 4.5 showing a higher soil loss for NT soils, for the first simulated storm event,
(significant) and slightly lower for the second storm (not significant)
64
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
Mean run-off
Run off (cm³)
2500
2000
1500
NT
1000
CT
500
0
Storm 1
Storm 2
Figure 4.6 Graph of the mean run-off
Fig 4.6, showing a higher run-off for the NT soils for the first simulated storm event, and
lower for the second (results for both the storm events are not significantly different).
Total infiltration
Infiltration (cm³)
3500
3400
3300
3200
3100
NT
CT
3000
2900
2800
2700
2600
Storm 1
Storm 2
Figure 4.7 Graph of total infiltration for the first and second
simulated storm event.
The infiltration for the CT soil (fig. 4.7) is significantly higher than for the NT soil; while for
the second storm event the NT soil had higher infiltration, though not significantly different.
For the first simulated storm event soil loss and run off were higher than expected for the
NT soil compared to the CT soil, while for the simulated storm event 2 the soil losses were
65
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
about the same (Figure 4.5 and Figure 4.6). Storm event 1 was statistically significant at
the 5% level, which is denoted in table 4.11 by different letters (i.e. a and b). Storm event
2 was not significantly different denoted by the same letter after the number. This would
probably be due to the aggregates being broken up and the clay particles clogging the pores
of the CT soil. Similarly the run-off was higher for the NT soil for the simulated storm event
1. Run-off was slightly lower for the NT soil compared to the CT soil. Both storms’ means
were not significant. This was probably to be expected as the ploughed soils were loosened
by the ploughing action.
In Table 4.11, the NT soils show a decrease in soil loss as well as a decrease in run-off,
while with the CT soils the reverse occurs. Total infiltration increases from the first to the
second storm event for the NT soils. The first storm event settles the soils and with the
second storm event the more stable soils give lower soil loss and run-off and higher
infiltration. The deduction that can be made is that the NT soils, due to their greater stability,
have a higher infiltration rate, resulting in the reduced run-off. The CT soils, due to the
reduced stability, have a reduced infiltration rate, possibly due to crust formation, leading to
increased run-off and soil loss.
The traditionally cultivated soils were loosened by the cultivation process, therefore it was
to be expected that the infiltration would be higher for storm event 1. For the simulated
storm event 2, the NT soil had a higher infiltration rate due to the soil structure being
maintained. Storm event 1 was significantly different. For the simulated storm event 2 the
infiltration was higher for the NT soil (Figure 4.7 & Table 4.11).
The average infiltration for each second rotation of the rainfall simulator, for the 18 NT and
CT soils, is given in Table 4.12; and Figures 4.8 and 4.9. The result illustrates the NT soils
remained stable over a longer period of time. Although the NT soils had a lower initial
infiltration rate for both simulated storm events, the CT soils’ infiltration rate decreased more
66
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
rapidly. The infiltration rate was significantly different for rotations 2 to 16 for the simulated
storm event 1. For simulated storm event 2, the NT soils had a significantly higher infiltration
rate for rotations 24 to 40 (Table 4.12).
Table 4.12 Average infiltration means for the eighteen NT and CT soils (n=72).
Simulated storm event 1
Simulated storm event 2
1
Rotation
NT
CT
p
Rotation
NT
CT
p1
2
186.2
211.5
0.001 a
2
209.8
213.5
0.162
4
180.8
206.3
0.001 a
4
205.4
207.4
0.397
6
174.5
201.7
0.001 a
6
197.4
198.6
0.609
8
170.5
198.3
0.001 a
8
191.8
193.6
0.548
10
165.3
190.4
0.001 a
10
186.3
191
0.236
12
160.3
182.5
0.001 a
12
180.6
185.1
0.387
14
154.7
172.5
0.004 a
14
176.7
178.4
0.77
16
148.3
163.87
0.022 a
16
170.8
170.5
0.966
18
141.1
153.5
0.11
18
165.2
160
0.499
20
133.8
145
0.176
20
158.7
148.3
0.208
22
127.9
135.6
0.358
22
151.1
137.2
0.103
24
120.8
126.5
0.493
24
144
125.8
0.036 a
26
112.5
118.9
0.441
26
135.9
114
0.011 a
28
105.8
111
0.524
28
129.7
105.4
0.005 a
30
98.6
104.1
0.476
30
122
97.9
0.004 a
32
92.9
94.5
0.822
32
114.6
92.1
0.006 a
34
86.6
85.9
0.916
34
108.4
85.6
0.004 a
36
80.9
80.3
0.917
36
101.2
79.4
0.006 a
38
76
75
0.886
38
94.8
76.9
0.02 a
40
72.1
71
0.897
40
89.6
73.6
0.035 a
42
67.7
70.1
0.701
42
83.2
71
0.105
44
65.8
69.2
0.599
44
81.1
69.9
0.134
46
65.8
69
0.615
46
80.1
69.9
0.172
48
65.8
69
0.615
48
80.1
69.9
0.172
50
65.8
69
0.615
50
80.1
69.9
0.172
The probability denoted by an “a” are significantly different.
1
p = probability value
67
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
Fig. 4.8 (storm 1) shows a gradual decrease in the rate of infiltration over time for both soils.
The rate of decrease is lower for the NT soils. The first 16 rotations were significantly higher
for the CT soil than the NT soils.
Average infiltration rate per rotation: Storm 1
250
Infiltration cm¯³)
200
150
NT
CT
100
50
0
2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50
Rotation
Figure 4.8 Infiltration means in the NT and CT for storm 1
The NT soils rate of infiltration for storm 2 is maintained at a higher rate for a longer period than
the CT soils. . The difference is statistically significant for rotations 24 to 40.
Average infiltration rate per rotation: Storm 2
250
Infiltration cm¯³)
200
150
NT
CT
100
50
0
2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50
Rotation
Figure 4.9 Infiltration means in the NT and CT soils for storm 2.
68
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
Figure 4.9 (storm 2) had a slightly lower initial infiltration rate, but decreased less than the
CT soil. There was a significantly different higher infiltration rate for the NT soils compared
to the CT soils for rotations 24 to 40.
Table 4.13 gives the average infiltration rates of the 72 samples measured in cm3, for the
NT and CT soils, per storm separately. The results show the initial infiltration rate (IIR) and
final infiltration rate (FIR) to be higher for the second storm event compared to the first
simulated storm event, for both the NT and CT farming systems.
Table 4.13 Average initial (IIR) (Rotation 2) and final infiltration rates (FIR) (Rotation
50) for the first and second simulated storms.
Difference
between
IIR and FIR
NT
IIR cm3
FIR cm3
First simulated
storm event
185.7
64.3
Second
simulated
storm event
204.7
77.9
% Difference
10.23
21.15
Difference
between IIR
and FIR
CT
IIR cm3
FIR cm3
121.4
212.9
70.7
142.2
126.7
213.1
71.5
141.5
0.16
1.13
The average initial infiltration- and final infiltration rate for the NT and CT treatments, in
Table 4.13, show there was a higher Initial Infiltration Rate (IIR) and Final Infiltration Rate
(FIR) for the second simulated storm event for both the NT and CT soils. The difference of
the IIR and FIR was higher for the NT soils at 10,23% and 21,15% respectively; compared
to 0.16% and 1.13% for the CT soils. The NT soils settled with the first storm event and
remained more stable. The difference between the IIR and FIR is lower for the NT treatment
compared to the CT treatment, for both simulated storm events. The difference between
the IIR and FIR is higher for the second storm than the first for the NT soil, compared to the
69
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
CT soil. This indicates that the CT soils slake more than the NT soils during the simulated
storm event.
Stern (1990) indicated that crust formation is a common phenomenon in arid and semi-arid
regions of the world, including South Africa. Seal formation was the dominant factor in
promoting water runoff and soil loss. If the crust is permanently formed, then it would be
expected that the initial infiltration rate for the second simulated storm would be lower. From
the results it would appear that if a crust was formed, it broke down as the soils dried out.
The soils in Mpumalanga are inherently stable due to the high iron and aluminium and low
smectite content. This could be the reason for the similar initial infiltration rates. The first
storm settled the soil resulting in the soils having a lower second storm initial infiltration rate
(personal communication Christal Bühmann, 2008). Literature cited showed results for one
simulated storm event only. This comparison for two simulated storm events has not been
reported in literature. The higher infiltration rate for the NT soils for the second simulated
storm event demonstrates the stability of the soils due to them not being worked, and the
benefits of the stabilising action of micro-organisms and fungi on aggregate formation and
stability.
4.3.4.2 Statistical Analysis – part 2
For the second part of the statistical analysis the farming systems (NT and CT) were
combined. The data was analysed as a split-plot design with the treatment, farming systems
(NT and CT) as main plots and the storm and rotations as sub-plots. Table 4.14 shows the
ANOVA for the combined farming systems (treatments). This dataset was analysed using
SAS 9.2.
70
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
Table 4.14 - ANOVA for the farming systems NT and CT (combined).
DF1
Source
Farm
Tmt
17
1
Error 1
MS2
p3
264549.0
86.8
17
74745.0
Storm
Rotation
TmtxRotation
TmtxStorm
StormxRotation
TmtxStormxRotation
Error 2
1
24
24
1
24
24
1666
207880.2
648713.2
6831.9
179310.3
394.7
477.8
3273.9
Total
7174
0.9732
<.0001
<.0001
0.0016
<.0001
1.0000
1.0000
1
Where DF = Degrees of freedom, 2MS = Means Square and 3p= Significant level of F-Ratio.
Tmt = treatment (NT and CT). A significant level less than 0.05 is considered as a significant
effect and indicated in bold.
Table 4.15 Average infiltration means for the NT and CT soils and different rotations.
Rotation
Tmt
N
2
2
4
4
6
6
8
8
10
10
12
12
14
14
16
16
18
18
20
20
22
22
24
NT
CT
NT
CT
NT
CT
NT
CT
NT
CT
NT
CT
NT
CT
NT
CT
NT
CT
NT
CT
NT
CT
NT
144
143
144
143
144
143
144
143
144
143
144
143
144
143
144
143
144
143
144
143
144
143
144
Mean
(cm3)
198
212.52
193.13
206.85
185.94
200.14
181.13
195.94
175.81
190.7
170.43
183.8
165.69
175.48
159.56
167.17
153.14
156.78
146.25
146.67
139.51
136.43
132.43
T
groupingA
bcd
a
cdef
ab
defg
abc
fgh
bcde
ghi
cdef
hij
efg
ijkl
ghi
jklm
ijk
lmn
klmn
no
mno
op
opq
pq
71
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
Table 4.15 Average infiltration means for the NT and CT soils and different rotations
(continued)
Rotation
Tmt
N
24
26
26
28
28
30
30
32
32
34
34
36
36
38
38
40
40
42
42
44
44
46
46
48
48
50
50
CT
NT
CT
NT
CT
NT
CT
NT
CT
NT
CT
NT
CT
NT
CT
NT
CT
NT
CT
NT
CT
NT
CT
NT
CT
NT
CT
143
144
143
144
143
144
143
144
143
144
143
144
143
144
143
144
143
144
143
144
143
144
143
144
143
144
143
Mean
(cm3)
126.17
124.22
116.43
117.74
108.18
110.31
101.02
103.76
93.29
97.5
85.71
91.06
79.86
85.4
76.01
80.83
72.45
75.46
70.56
73.47
69.51
72.99
69.44
72.99
69.44
72.99
69.44
TgroupingA
qr
qr
rst
rs
stu
stu
uv
tuv
vwx
uvw
wxyz
vwxy
AByz
Awxyz
ABz
ABxyz
AB
ABz
B
ABz
B
ABz
B
ABz
B
ABz
B
LSD(p=0.05) = 13.249
Means within each column with the same letter or letters do not differ significantly at the 5%
significant level.
A
T-grouping where capital letters might occur, it is different to lower case letters
Table 4.15 and figure 4.10 give the means for each treatment (NT and CT), rotation
interaction. Looking at it, it can be seen that the treatment CT for rotations 2 to 12, for both
storms, differs statistically (better infiltration) for the same rotations, from the NT at a 5%
significance level. At later rotations (22 to 50), the NT treatment performs better (better
infiltration) - although not statistically different.
72
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
Average infiltration rate per treatment
Average infiltration (cm3
250
200
LSD(p=0.05)=13.249
150
NT
CT
100
50
0
2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50
Rotation
Figure 4.10 Average infiltration rate for the 2 storms for each rotation.
Table 4.16 and figure 4.11 show the means for the combined rotations for each treatment(NT, CT) and storm interactions. Within storm 1 the treatment CT did statistically better than
the CT treatment over all the rotations. Within storm 2 the pattern changed; the NT treatment
performed statistically better. The infiltration for the treatment CT did not change over the 2
storms. As expected partly due to the loosening of the soil by the ploughing action, the
infiltration rate for the first storm was significantly higher. The NT soil had an improved
infiltration rate for the second storm, due to the settling effect of the first storm. This would
be attributed to the soil maintaining its’ structure for a longer period of time.
Table 4.16 Average infiltration rate for the 2 storms combined as per treatment (NT
and CT).
Storm
1
1
2
2
Tmt
NT
CT
NT
CT
N
1800
1775
1800
1800
Mean
(cm3)
116.82
127.01
137.55
127.39
Tgrouping
c
b
a
b
LSD(p=0.05) =3.7475
Means within each column with the same letter or letters do not differ significantly at the 5%
significant level
73
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
Infiltration means of treatments versus storm
140
Infiltration (cm¯³)
135
LSD(p=0.05)=3.74
130
NT
CT
125
120
115
110
105
1
Storm
2
Figure 4.11 Average infiltration means as per treatment (NT and CT).
74
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
CHAPTER 5
CONCLUSIONS AND RECOMMENDATIONS FOR FURTHER STUDY
LandCare is a community-based programme with the overall goal of optimizing productivity
through the sustainable use of natural resources, resulting in greater food security, job
creation and better quality of life for all. A community-driven approach was followed with
the core principles being the training and empowerment of land users and community
members in the principles of conservation agriculture technologies. Training was in the
form of farmer managed research demonstrations and farmer demonstration trials were the
thrust with the Mlondozi project. Eighteen farmers initially joined the programme.
Mlondozi is a low income area which is also far from the main centres, thus making farming
difficult and costly. The LandCare approach was a successful programme in that yields
increased with reduced labour. With the farmers planting the NT alongside the CT plot, they
were able to compare the two production systems. This resulted in the farmers taking
ownership of the experimentation on their respective farms. The enthusiasm was such that
the conventional production plots, which were meant to be cultivated following the
conventional practices for the duration of the study period, for the farms and researchers to
monitor and compare the differences were, over the study period, converted to NT.
Due to the granitic parent material, the soils are inherently acidic, which is the second
largest limiting factor in maize production after rainfall, in Mlondozi. With an increase in soil
acidity there was a decrease in Ca and Mg and an increase in Al. The recommended 5 ton
of lime ha-1 proved to be beneficial as the soils’ pH and nutrient status gradually improved
for the basic elements during the study period. The fertility levels for Ca, Mg and P of the
top soil for the NT farming system gradually increased over the years, and the pH also had
an upward trend over the study period (Table 4.3).
75
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
The aim of the project was to optimise the production of maize by introducing an improved
farming system (NT) by planting hybrid seed, applying lime to correct the pH, correct
fertilizer type and quantity applied, earlier planting and weed control by the use of
herbicides. The yield of the NT system was higher than the CT systems over the study
period. The 2001-2002 season was a below average rainfall year, accompanied by hail,
resulting in reduced yield, compared to the previous years. The improved yields meant the
farmers were able to grow sufficient maize for their own consumption and sell the excess
to buy inputs, thus improving their standard of living. The improved farming system led to
higher yields and improved drought tolerance, thus ensuring food security.
A higher root density is required for optimum nutrient and water uptake as 80% of the feeder
roots occur in the top 100 mm of the soil. The root density for NT in the top- and sub-soil
increased from 1999 to 2001 but decreased in 2002 due to a lower than normal rainfall
season. By keeping the soil covered a more favourable condition is created with regard to
temperature and soil water content, for root growth.
The FAO (2001, 2007) definition of NT includes keeping the soil covered by a mulch or
cover crop. Retaining the crop residue aids in maintaining soil water content by reducing
evaporation and improving infiltration. Rain drop impact was adsorbed by the mulch, thus
reducing surface crusting. A more conducive condition was created for earthworm activity.
Mulch cover increased from 30% to 60% for 2000 and 55 % for the remaining years (2001
and 2002). An observation the farmers repeatedly made for themselves was the way the
NT maize plants continued to grow unhindered between rain fall events, while the
conventionally cultivated maize would be wilting within a few days to a week after a storm.
76
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
Mlondozi is a high rainfall area, which could result in soil erosion taking place on the
conventionally cultivated soils (CT). At the end of the project four undisturbed soil samples
were taken from both the CT and NT fields of the eighteen farmers who initially joined the
training programme. The samples were placed in a laboratory rainfall simulator for two
successive storm events where the run-off, infiltration and soil loss was measured. From
the results above, the implementation of the NT farming system resulted in a reduced runoff, improved infiltration and reduced soil erosion. Following the conventional cultivation
practice the fields were ploughed yearly for weed control and seed bed preparation. In so
doing aggregates were broken along with the fungal hyphae and organic gels, which would
otherwise have maintained the soil’s stability.
The CT soils disturbance by the regular cultivation resulted in a reduction in the infiltration
rate compared to the NT soils. The NT soils improved stability resulted in a lower reduction
in infiltration rate that could be attributed to the fungal hyphae which were not damaged by
regular cultivation, adding to aggregate stability. The Conventional tillage soils had a lower
soil loss for the first storm. The soil loss for both the NT and CT soils was the same for the
second storm. The “No Till” soils showed reduced run-off for the two storm events and the
infiltration increased. The CT soils had a similar infiltration rate for the two rainfall events.
From the above results it is recommended that NT be implemented for sustainable crop
production.
The pan method:
Traditionally soil samples taken in the field were dried; clods pulverised then sieved thus
breaking the aggregates. The pan method was as an improvement in that undisturbed soil
samples were taken.
•
The soil samples taken using the pan method showed consistent similarity of the
results for the four replicates for each treatment. Further investigation could be
77
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
undertaken, to take the soil samples using a perforated metal tray, which could be
placed directly into the laboratory rainfall simulator, resulting in even less soil
disturbance.
•
Soil samples taken using the pan method gave results quicker than would have been
the case with the other methods. Results could be obtained within 2 hours after
returning to the laboratory, compared to months for in-field run-off plots.
Further study and recommendations
No Till is a modern day agricultural revolution. There is no other farming practice that has
had such an impact on sustainable crop production. This study paves the way for further
research on the effects of No Till versus Conventional agricultural practice:
•
It is recommended that a planned trial be laid out (NT versus CT) with the following
treatment applications:
•
Plough versus no-tillage,
•
Fertiliser,
•
Liming;
•
Weeding.
These treatments could be tested against one another for soil erosion, run-off
and infiltration.
•
To further verify the stability of the NT soils it is recommended that samples be
placed in the rainfall simulator for a third and possibly fourth storm event to further
verify the stability of the “No-Till” soils.
•
Further research needs to be conducted to investigate crusting, clay mineralogy and
the effects of sodium and magnesium on the stability of aggregates for NT and CT
soils under different rainfall conditions. The soils used in the study had a high
kaolinite content (31% to 70%) which explains the high stability along with the fact
78
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
that they are highly weathered. Aggasi, Shainberg & Morin (1988) found that there
was a partially formed crust with a short duration rain event or low intensity long
duration storm event. It can be assumed that a partially formed crust formed would
result in the infiltration rate being the same at the start of the second storm event,
but this needs to be verified.
79
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
REFERENCES
ABIVEN, S., MENASSERI, S. AND CHENU, C. 2009. The effects of organic inputs over time
on soil aggregate stability – A literature analysis. Soil Biology & Biochemistry 41: 1-12.
ACOCKS, J.P.H. 1988. Veld types of Southern Africa (3rd ed.) Memoirs of the Botanical Society
of Southern Africa 57:1-146.
AGGASI, M., SHAINBERG, I. AND MORIN, J. 1981. Effect of electrolyte concentration and
soil sodicity on infiltration rate and crust formation. Soil Sci. Soc. Am. J. 45: 848-851.
AGGASI, M., SHAINBERG, I. AND MORIN, J. 1988. Effects on seal properties of changes in
drops energy and water salinity during a continuous rainstorm. Aust. J. Soil Res. 26, 651659.
AGROMET 2010. Agromet database, ARC-Institute for Soil Climate and Water, Pretoria.
ALBERTS, L. 1986. 1:250000 2630 Mbabane, Geological map. Geological survey, Pretoria,
South Africa.
ANGERS, D.A. AND CHENU, C. 1998. Dynamics of soil aggregation and C sequestration. By:
LAL, R., KIMBLE, R.M., FOLLET, R.F. AND STEWARD, B.A. eds. 1998. Soil Processes
and the carbon cycle. CRC Press LLC, NY, USA.
AON, M.A., SARENA, D.E., BURGOS, J.L. AND CORTASSA, S. 2001 (Micro)biological,
chemical and physical properties of soils subject to conventional and no-till management:
An assessment of their quality status. Soil & Tillage res. 60. 173 – 186.
ARC-ISCW. 2004. Overview of the status of the agricultural natural resources of South Africa.
ARC-ISCW Report No. GW/A/2004/13, ARC-Institute for Soil, Climate and Water, Pretoria.
ARC-ISCW. 2007 Operation and maintenance of the rainfall simulator. QMS document
number: ISCW/SWS/PRC/006
BARNARD, R.O., VAN DER MERWE, A.J., DE VILLIERS, M.C., VAN DER MERWE, G.M.E.
and MULIBANA, N.E., 2000. Problem soils including degraded soils: Extent, present use,
management and rehabilitation in South Africa. Proc. FAO/ISCW Expert Consultation on
Management of Degraded Soils in southern and east Africa (MADS-SEA) 126-140. Pretoria.
80
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
BATIONO, A. AND BUERKERT, A. 2001. Sol organic carbon management for sustainable
land use in Sudano-Sahelian West Africa. Nutrient Cycling in Agroecosystems 61: 131-142.
BEARE, M.H., HENDRIX, P.F. and COLEMAN, D.C. 1994. Water stable aggregates and
organic matter fractions in conventional- and no-tillage soils. Soil Sci. Soc. Am. J. 58: 777786.
BLOEM, A.A. 1992. Criteria for adaption of the design and management of overhead irrigation
systems to the infiltrability of soils. M.Sc. Univ. Pretoria.
BOOYENS, H.B., POTGIETER, L.J.C. & MATLAWA, S.M. 2000. Characterisation of soils in
the Mlondozi district of Mpumalanga. ISCW Report number: GW/A/2000/71. ARC - Institute
for Soil, Climate and Water, Pretoria, S. A.
BOT, A. AND BENITS, J. 2005. The importance of soil organic matter: Key to drought-resistant
soil and sustained food production. FAO soils bulletin 80, Food and Agriculture Organization
of the United Nations, Rome, Italy.
BOX, J.E. AND BRUCE, R.R. 1996. The effect of surface cover on infiltration and soil erosion.
In: M. AGASSI (ed), Soil erosion, Conservation and Rehabilitation. Marcel Dekker, NY.
BRADFORD, J.M. AND HUANG, C. 1996. Splash and Detachment by Water drops. AGASSI,
M. Soil erosion, Conservation and Rehabilitation. Marcel Dekker, NY.
BRAY, R.H AND KURTZ, L.T. 1945. Determination of total, organic and available forms of
phosphorus in soils. Soil Sci. 59, 39-45.
BROWN, S. M., WHITWELL, T., TOUCHTON, J.T. and BURMESTER, C.H. 1985.
Conservation tillage for cotton production. Soil Sci. Soc. Am. J. 49, 1256-1260.
CHIVENGE, P.P., MURWIRA, H.K., GILLER, K.E., MAPFUMO, P. AND SIX, J. 2007. Longterm impact of reduced tillage and residue management on soil carbon stabilization:
Implications for conservation agriculture on contrasting soils. Soil & Tillage res. 94: 328337.
CLAASSENS, A.S. and VAN DER WATT, H.v.H. 1993. An inexpensive, portable rain
simulator: construction and test data. S.A.J. Plant and Soil 10 (1).
81
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
DEPARTMENT OF WATER AFFAIRS. 1986. Management of the water resources of the
Republic of South Africa. Dept. Water Affairs, Pretoria, SA.
DERPSCH, R. 2008. No-tillage and conservation agriculture: A progress report. In: T.
GODDARD, M. ZOEBISCH, Y. GAN, W. ELLIS, A. WATSON AND S. SOMBATPANIT
(eds.). No-till farming systems. Special Publication No. 3, World Association of Soil and
Water Conservation, Bangkok, ISBN: 978-974-8391-60-1. (P 33)
D’HAESE, L. AND KIRSTEN, J. 2006. Rural development focusing on small scale agriculture
in Southern Africa. Univ. Pretoria. S. A.
D’HUYVETTER, J.H.H. 1985. Determination of threshold slope percentage for the
identification and delineation of arable land in Ciskei. M.Sc. Agric., Univ. Fort Hare.
DIALLO, D., BOLI, Z. and ROOSE, E. 2008. Influence of no-tillage on soil conservation, carbon
sequestration and yield of intensive rotation maize-cotton: research on sandy Alfisols of
Cameroon and Mali. In: GODDARD, T., ZOEBISCH, M., GAN, Y., ELLIS, W., WATSON, A.
AND SOMBATPANIT, S. (eds) 2008. No-till farming systems. Special Publication No. 3,
World Association of Soil and Water Conservation, Bangkok, ISBN: 978-974-8391-60-1.
ECK, K.J and BROWN, D.E. 2004. Estimating corn and soybean residue cover.
http://www.extension.purdue.edu/extmedia/AY/AY-269-W.pdf. Accessed 29/10/2010.
FAO, 2001. Conservation Agriculture: Case studies in Latin America and Africa. FAO soils
Bulletin 78.
FAO. 2007. Agriculture and Consumer Protection Department: Conservation Agriculture.
Online www.fao.org/ag/ca Accessed April 2007.
GARLAND, G., HOFFMAN, T. & TODD, S., 2000. Chapter 6: Soil degradation. In: Land
degradation in South Africa. T. Hoffman, S. Todd, Z. Ntshona & S. Turner (Eds). Dept. Env.
Affairs and Tourism, Pretoria.
GEOLOGICAL SURVEY, 1970. Geological map of the Republic of South Africa and Kingdoms
of Lesotho and Swaziland. (1:1 000 000) Geological Survey S. Afri, Pretoria
GEOLOGICAL SURVEY 1984. 1 250 000 Geological series 2530 Barberton. Geological
Survey, Pretoria, SA.
82
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
GODDARD, T., ZOEBISCH, M., GAN, Y., ELLIS, W., WATSON, A. AND SOMBATPANIT, S.
(eds) 2008. No-till farming systems. Special Publication No. 3, World Association of Soil
and Water Conservation, Bangkok, ISBN: 978-974-8391-60-1.
GOLDBERG, S. and GLAUBIG, R.A. 1987. Effect of saturating cation, pH and aluminium and
iron oxides on the flocculation of kaolinite and montmorillonite. Clays Clay miner. 35:220227.
GUGINO, B.K., IDOWU, O.J., SCHINDELBECK, R.R., VAN ELS, H.M.,
WOLF, D.W.,
MOEBIUS-CLUNE, B.N., THIES, J.E. AND ABAWI, G.S. 2009. Cornell soil health
assessment training manual, Ed 2.0, Cornell Univ., Geneva, NY.
HENSLEY, M. and LAKER, M.C. 1975. Land resources of the consolidated Ciskei. In: M.C.
Laker (ed.) The Agricultural Potential of the Ciskei, Univ. Fort Hare, Alice.
HOROWITZ, J., EBEL, R. and UEDA, K. 2010. “No-Till” Farming Is a Growing Practice. USDA,
Economic Information Bulletin No. 70.
HUSSON, O. 2003. Soils as dynamic biological systems: Methods for enhancing soil fertility
and restoring 'problem' soils. Soil health and remediation workshop, Pietermaritzburg,
Natal, S. A.
JANSEN VAN RENSBURG, H.G. (ed.) 2001. Progress report for the Mlondozi LandCare
Project. ARC – ISCW, Pretoria. S. A. Report Number GW/A/2001/60.
JANSEN VAN RENSBURG, H.G. (ed.) 2002. Final Report for Mlondozi LandCare Project.
ARC –ISCW, Pretoria, S. A. Report number GW/A/2003/05
JANSEN VAN RENSBURG, H.G. 2002. Demonstration and assessment of sustainable land
management practices in the Mlondozi district of Mpumalanga. ARC- Institute for Soil
Climate and Water, Pretoria S. A. Report Number GW/A/2003/05.
JANSEN VAN RENSBURG, H.G. AND MLONDOZI PROJECT TEAM. 2001 Mlondozi
LandCare: Training needs survey. ARC – ISCW, Pretoria, S. A. Report number
GW/A/2001/38.
83
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
JANSEN VAN RENSBURG, H.G., ENGELBRECHT, H. AND SMITH, H.J. 2003. Mlondozi
LandCare Project: Impact Assessment. ARC – ISCW, Pretoria, S. A. Report number
GW/A/2003/49.
JANSEN VAN RENSBURG, H.G. 2009. Soil chemical and nutrient uptake dynamics of maize
(Zea mays L.) as affected by neutralization and re-acidification after liming. Ph. D. Thesis
Pretoria Univ.
JASTROW, J.D. AND MILLER, R.M. 1998. Soil Aggregate stabilisation and carbon
sequestration: Feedbacks through organomineral associations. By: LAL, R., KIMBLE, R.M.,
FOLLET, R.F. AND STEWARD, B.A. eds. Soil Processes and the carbon cycle. CRC Press
LLC, NY, USA.
JEHNE, W. 1981. Endomycorrhizae in plant colonization on coastal sand-dues at Cooloola
Queensland. Australian. J of Ecol. 6: 221-230.
JEZILE, G.G. 2004. Cattle manure as a fertilizer. (Un-published), ARC – Institute for Soil,
Climate and Water, Pretoria.
JIN, K., CORNELIUS, W.M., GABRIELS, D., SCHIETTECATTE, W., DE NEVE, S., LU, J.,
BUYSSE, T., WU, H., CAI, D., JIN, J. AND HARMANN, R. 2008. Soil management effects
on runoff and soil loss from field rainfall simulation. Catena 75, 191-199.
JONES, D.L., NGUYEN, C. AND FINDLEY R.D. 2009 Carbon flow in the Rhizosphere: carbon
trading at the soil – root interface. Plant and Soil 321: 5-33.
KaNgwane (South Africa homeland). http://flagspot.net/flags/za-ng.html Accessed 4/10/2010.
KENT, L.E. 1980. Stratigraphy of South Africa; Part 1: Lithostratigraphy of the republic of South
Africa, South West Africa/Namibia and the Republics of Bophuthatswana, Transkei and
Venda. Handbook. Geological Survey 8, Pretoria, South Africa.
KLADIVIO, E.J. 2001. Tillage systems and soil ecology. Soil & Tillage Research 61; 61-76.
KLUCAS, G. 1999. Australian LandCare: what can we learn? Conservation voices, Australia
KONGO V.M. AND JEWITT, G.P.W. 2006. Preliminary investigation of catchment hydrology
in response to agricultural water use innovations: A case study of the Potshini catchment –
South Africa. Physics and chemistry of the earth 31: 976-987.
84
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
KUNDU, S., BHATTACHARYYA, R., PRAKASH, V., GOSH B.N. and GUPTA, H.S. 2007.
Carbon sequestration and relationship between carbon addition and storage under rain-fed
soybean-wheat rotation in a sandy loam soil of the Indian Himalayas. Soil & Tillage
Research 92:87-95.
LAIRD, D.A. 2001. Nature of clay-humic complexes in an agricultural soil: II. Scanning electron
microscopy analysis. Soil Sci. Soc. Am. J. 65:1419-1425.
LAKER, M.C. 2004 Advances in soil erosion, soil conservation, land suitability, evaluation and
land use planning research in South Africa, 1978-2003. S. A. J. Plant Soil 21 (5)
LAKER, M.C. 2012. Appropriate plant nutrient management for sustainable agriculture in
Southern Arica. Unpublished, Pretoria University, South Africa.
LEVY, G.J. 1988. The effects of clay mineralogy and exchangeable cations on some of the
hydraulic properties of soils. D Sc. (Agric) Univ. Pretoria
LOCH, R.J., CONNELLY, R.D. and LITTLEBOY, M. 2000. Using rainfall simulation to guide
planning and management of rehabilitation areas: Part II. Computer simulations using
parameters from rainfall simulation. Land Degrad. Dev. 11, 241-255.
LOCH, R.J., ROBOTHAM, B.G., ZELLER, L., MASTERTON, N., ORANGE, D.N., BRIDGE,
B.J., SHERIDAN, G. AND BOURKE, J.J. 2001. A multi purpose rainfall simulator for field
infiltration and erosion studies. Aust. J. Res. 39(599-610).
MCHUGH, A.D., TULLBERG, J.N. AND FREEBAIRN, D.M. 2009. Controlled traffic farming
restores soils structure. Soil & Tillage Res. 104. 164-172.
MIDGLEY, 1952. A preliminary survey of the surface water resources of the Union of South
Africa. Unpublished PhD thesis. University of Natal, Pietermaritzburg
MILLER, W.P. 1987. A solenoid-operated, variable intensity rainfall simulator. Soil Sci. Soc.
Am. J. 51, 832-834.
MOKMA, D.L. AND SIETZ, M.A. 1992. Effects of soil erosion on corn yields on Marlette soils
in south-central ichigan. J. of Soil and Water Conservation 47(4): 325-327.
85
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
MOLOPE, M.B. 1999. Implementation framework for the LandCare programme, Department
of Agriculture and forestry, Resource Conservation and quality Control, Pretoria.
www.daff.gov.za/docs/Landcare/landcare.htm accessed 20/02/2012.
MORIN, J., BENYAMINI, Y. AND MICHAELI, A. 1981. The dynamics of soil crusting by rainfall
impact and water movement in the soil profile. J. Hydrol. 52: 321-335.
MORIN, J., GOLDBERG, S. and SEGINER, I. 1967. A rainfall simulator with a rotating disc.
Trans. Am. Soc. Agric. Engrs. 10, 74-79.
MULIBANA, N.E. 2001. Analysis and synthesis of data on relationships between soil factors
and soil erosion in South Africa. M.Inst.Agrar. dissertation, Univ. Pretoria.
MUNN, J.R. and HUNTINGTON, G.L. 1976. A portable rainfall simulator for erodability and
infiltration measurements on rugged terrain. Soil Sci. Soc. Am. J. 40, 622-624.
MYBURG, W.J. AND BREYTENBACH, P.J.J. 2001. Rangeland description and evaluation of
the Mlondozi LandCare site for livestock production. ARC-Range and Forage Institute,
Pretoria.
No-Till, Advantages and benefits in crop production. 2005. No-till club of KwaZulu-Natal.
NYAKATAWA, E.Z., REDDY, K.C. and LEMUNYON, J.L. 2001. Predicting soil loss in
conservation tillage cotton production systems using the revised universal soil loss equation
(RUSLE). Soil and Till. Res. (57) 213-224.
PAYNE, R.W., MURRAY, D.A., HARDING, S.A., BAIRD, D.B. & SOUTER, D.M. 2007. GenStat
for Windows® (10th Edition) Introduction. VSN International, Hemel Hempstead, UK.
RAPP, I. 1998. Effects of soil properties and experimentation conditions on the rill erodabilities
of selected soils. Ph.D. thesis. Univ. Pretoria.
Revised Universal Soil Loss Equation (RUSLE) 1.02. 1993. Soil and Water Conservation
Society, U. of Tennessee, USA
ROOSEBOOM, 1976. Reservoir sediment deposition rates. Proceedings of the Second
International Congress on Large Dams, Mexico.
ROTH, H., MEYER, B. and FREDE, G. 1985. A portable rainfall simulator for studying factors
affecting runoff, infiltration and soil loss. Catena 12 (79-85)
86
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
RUKUNI, M. 1998. “Why Land tenure policy is central to Africa’s future governance, economic
and social progress?” Mini symposium paper prepared for the seminar on Leadership for
transformation and organisational change in African agriculture, Victoria Falls, 21
September to 9 October 1998
RUTHERFORD, M.C. and WESTFALL, R.H. 1986. Biomes of southern Africa - An objective
categorization. Memoirs of the Botanical Survey of South Africa 54:1-98.
SAS Institute, Inc. (1999), SAS/STAT User's Guide, Version 9.2, Volume 2. SAS Institute Inc,
SAS Campus Drive, Cary, North Carolina 27513.
SCHROEDER, D. 1984. Soils: Facts and Concepts. Int. Potash Institute, Bern, Switzerland.
SCHWARTZ and PULLEN, 1996. A guide to the estimation of sediment yield in South Africa.
Civil Engineering in South Africa. December 1966, 343-346
SCOTT, D.F. AND VAN WYK, D.B. 1992. The effects of fire on soil water repellency,
catchment sediment yields and streamflow. In: van Wilgen, B.W., Richardson, D.M., Kriger,
F.J. and van Hensbergen, H.J. (eds) Fire in South Africa Mountain Fynbos: ecosystem,
community and species response at Swartboskloof. Springer-Verlag, Berlin p 217-239.
SEEGER, M. 2007. Uncertainty factors determining runoff and erosion processes as quantified
by rainfall simulations. Catena 71:56-67.
SHAINBERG, I AND LEVY, G.J. 1996. Infiltration and seal formation process. In Soil erosion,
conservation and rehabilitation Aggasi, M. (ed.), Marcel Decker, Inc, NY.
SHELTON, D.P AND JASA, P.J. 1998. Estimating percent residue cover using the photocomparison method, G95-1134-A. www.ianr.unl.edu/PUBS/fieldcrops/g1134.htm
accessed 08/01/2003.
SILBURN, D.M., HARGREAVES, P., BUDD, N. and GRANVILLE, S.G. 1996. Endosulfan on
cotton plants – persistence and wash-off during rain. In ‘INTERSECT 96 – International
SILBURN, D.M., WATERS, D.K., CONNOLLY, R.D., SIMPSON, B.W. and KENNEDY, I.R.
1998. Techniques for stabilising soil erosion on cotton farms. In ‘Minimising the impact of
pesticides on the riverine environment: key findings from research with cotton industry ‘.
87
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
(Eds N.J. SCHOFIELD, V.E. EDGE) Occasional Paper 23/98, pp 99-105. (Land and Water
Research and Development Corporation: Canberra)
SMITH, H.J.C. 1990. The crusting of red soils as affected by parent material, rainfall, cultivation
and sodicity. M.Sc. (Agric), Univ. Pretoria.
SNEDECOR, GW & COCHRAN, WG. 1980. Statistical methods (7th Ed.). Iowa State University
Press.Symposium on Pesticide Management Catchments’. 4-5 Feb. 1998. University of
Southern Queensland, Toowoomba. Pp107-117. (Condamine-Balonne Water Committee:
Dalby, Qld)
SO, H.B., GRABSKI, A. and DESBOROUGH, P. 2009. The impact of 14 years of conventional
and no-till cultivation on the physical properties and crop yields of a loam soil at Grafton
NSW, Australia. Soil & Tillage res. 104. 180 – 184.
SOIL CLASSIFICATION WORKING GROUP, 1991. Soil classification: A Taxonomic system
for South Africa. Department of Agricultural Development, Pretoria, SA.
STERN, R., 1990. Effects of soil properties and chemical ameliorants on seal formation, runoff
and erosion. D.Sc. (Agric), Univ. Pretoria.
STEVENSON, F.J. 1994. (2nd ed). Humus chemistry: Genesis, Composition, reactions. John
Wiley & sons, NY. USA.
STUBBS, T.L., KENNEDY, A.C. and SCHILLINGER, W.F. 2004. Soil ecosystem changes
during the transition to no-till cropping. http://www.haworthpress.com/web/JCRIP, accessed
8 Oct. 2010.
SUMNER, M.E., 1957. The physical and chemical properties of Tall Grass Veld soils of Natal
in relation to their erodability. M.Sc. (Agric.), Univ. Natal.
SWANSON, N.P. 1965. A rotating-boom rainfall simulator. Trans. of the ASAE 8, 71-72.
TANG, Y., ZANG, H., SCHRODER, J.L., PAYTON M.E. and ZHOU, D. 2007. Animal Manure
reduces aluminium toxicity in acid soil. SSSAJ 71: 6.
THE NON-AFFILIATED SOIL ANALYSIS WORK COMMITTEE, 1990. Handbook of standard
soil testing methods for advisory purposes, Report: ISBN 0-620-14800-4. Soil Science
Society of South Africa, Pretoria.
88
© University of Pretoria
University of Pretoria – Kidson, M.V. (2014)
TISDALL, J.M. and OADES, J.M. 1982. Organic matter and water-stable aggregates in soils.
J. of Soil Sci. 33: 141-163.
TISDALL, J.M., NELSON, S.E., WILKINSON, K.G., SMITH, S.E., and MCKENZIE, B.M. 2012.
Stabilisation of soil against wind erosion by six saprotrophic fungi. Soil Biol. & Biochemistry
50: 134-141.
TURNER, D.P. 2004. Soils of KwaZulu-Natal and Mpumalanga: Recognition of Natural soil
bodies. Ph.D. Thesis Pretoria Univ.
VAN DER MERWE, A.J., DE VILLIERS, M.C., BARNARD, R.O., BEUKES, D.J., LAKER, M.C.,
& BERRY, W.A.J., 2000. Technical report on guidelines on the management and
rehabilitation of acid and fertility declined soils. Proc. FAO/ISCW Expert Consultation on
Management of Degraded Soils in southern and east Africa (MADS-SEA). 141-171.
Pretoria.
VIDEO: 2013. The next step, adding cover crop to a no-till system.flv. WWW. SARE.org.
Accessed 25 February 2013.
WALKLEY, A.S. AND BLACK, I.A. 1934. An examination of the set-java method for
determining soil organic matter and proposed modification of the chromic acid titration
method. Soil Sci. 37, 29-38.
WEAVER, A. and VAN BREDA. 1991. The distribution of soil erosion as a function of slope
aspect and parent material in Ciskei, Southern Africa. GeoJournal 23.1 29-34.
WESSELS, J.G.H. 1996. Fungal hydrophobians: Proteins that function at an interface. Trends
Plant Sci. 1:9-15.
WILSON, G.V., McGREGOR, K.C. AND BOYKIN, D. 2008. Residue impacts on runoff and
soil erosion for different corn plant populations. S Soil & Tillage Res. 99. 300-307.
XABA, G.O. 2002. An investigation on how information campaign on liming and conservation
tillage was managed and received in the Mlondozi LandCare project. MA, Mini dissertation,
University of Pretoria, Department of Information Science, Pretoria.
89
© University of Pretoria
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertisement