natural resources systems programme - gov.uk · dry matter accumulation results of dry matter...
TRANSCRIPT
NATURAL RESOURCES SYSTEMS PROGRAMME PROJECT REPORT1
DFID Project Number
R8088A
Report Title
Using the PT model to determine appropriate management for maize production. A case study of Fulwe Village in Morogoro District for Morogoro District.
Annex B19[2] of the Final Technical Report of project R8088A.
Report Authors
Pangapanga, C.P., Marijani, B.J., Tumbo, S.D., Mapulila, T. and Rwehumbiza, F.B.
Organisation
SWMRG, Sokoine University of Agriculture, Tanzania
Date2005
NRSP Production System Semi Arid
1 This document is an output from projects funded by the UK Department for International Development (DFID) for the benefit of developing countries. The views expressed are not necessarily those of DFID.
Case Study Two
Morogoro District – Fulwe Village
Using the PT Model to Determine Appropriate WaterManagement for Maize Production: A Case Study of
Fulwe Village in Morogoro District
C. P. Pangapanga1, B. J. Marijani2, S. D. Tumbo3, T. Mpulila3, F. B. Rwehumbiza3
1Agricultural and Livestock Development Office, Morogoro District Council 2Agricultural and Livestock Development Office, Mvomero District Council
3Soil-Water Management Research Group, Sokoine University of Agriculture
Background
Fulwe village is located in Mikese ward, some 30 kilometres east of the Morogoro
Municipality, along the Morogoro - Dar es Salaam highway. It is located at longitude
37º53’60’’ East, and the latitude is 6º46’ south. The annual rainfall ranges from 700mm
to 1000 mm. The short rainy season (Vuli) starts in mid-October and ends in December
while the long rains (Masika) start in February and ends in mid May. The dry season
extends from June to October. The annual average maximum and minimum temperatures
are 26ºC and 21ºC respectively. Soils are acidic lithosols and ferralitic latosols with
deeper deposits of ferruginous sandy clay. Fulwe villagers depend mostly on rainfed
agriculture and maize is the main staple food and also one of the major sources of
income. Other crops cultivated include beans, soybeans and horticultural crops.
Farmers in the village have been facing a problem of good and poor maize yields even in
the same season. The reasons for the mixed results were not well known by the
agricultural extension officers serving the area. Some of the reasons that were being cited
include inadequate crop husbandry, non-use of certified maize seeds and improper
planting dates. The agricultural extension officers thought the problems were related to
improper planting dates. That is the variable yields among farmers were caused by some
farmers planting at the start of the rainy season and some in the middle and others very
late in the rain season. It was noted from the agricultural extension officers that they were
unable to successfully advise farmers on proper planting dates. In some cases they
advised farmers to plant once the rain started, but still farmers ended up with lower yields
PT Model Help Office 14
Sokoine University of Agriculture
Case Study Two
Morogoro District – Fulwe Village
compared to those who planted a bit late. The main issue that confronted agricultural
extension officers, was therefore on finding tools or methods that will assist them in
identifying a proper planting window for farmers in Fulwe village.
It is against this background that the PT Model, which is an agro-hydrological software,
was therefore identified by the agricultural extension officers as a tool that could assist in
the identification of the proper planting window by simulating maize yields for different
planting dates in Fulwe village.
Description of the PT Model
PARCHED-THIRST stands for Predicting Arable Resource Capture in Hostile
Environments During the Harvesting of Incident Rainfall in the Semi-arid Tropics.
PARCHED-THIRST model is a user-friendly, process-based model, which combines the
simulation of hydrology with growth and yield of a crop on any number of distinct or
indistinct runoff producing areas (RPAs) and runoff receiving areas (RRAs). It is a
distributed model, which simulates the rainfall-runoff process, soil moisture movement
and the growth of sorghum, rice, maize and millet in response to daily climate data. The
model includes the simulation of rice and macro-catchment systems up to the hillside or
small catchment scale. The schematic diagram of its components is shown in Figure 1.
The model allows the user to enter dates for the start of the seasons (Vuli and/or Masika)
and for the sowing dates. Alternatively the software itself can determine the best planting
date. Furthermore, the software provides daily outputs of soil-moisture values for
different layers of soils, dry matter accumulation, grain filling and other variables.
PT Model Help Office 15
Sokoine University of Agriculture
Case Study Two
Morogoro District – Fulwe Village
Figure 1: Interaction between the components making up the PARCHED-THIRST
model.
DATA PRE-PROCESSORS
USER
User
ClimateGenerator
RainfallDisaggregator
PedotransferFunctions
RunoffRouting
InfiltrationModel
RUNOFF MODEL
ModifiedORYZA_W
ModifiedPARCH
CROP MODEL
Daily weather
Rainfall intensity
Soil waterparameters Modified
LOWBAL
Piston Flow
Fl
SOIL WATER MODEL
FiniteDifference
Output crop and hydrological data
Input parametersand data
Objectives of the study
The main objective of the study was, to investigate the reasons for poor maize yield and
determine the appropriate management options that can improve maize production in
Fulwe village, Morogoro district.
The specific objectives included the following:
To investigate the effect of planting dates on maize yield in Fulwe village using the
PT Model.
PT Model Help Office 16
Sokoine University of Agriculture
Case Study Two
Morogoro District – Fulwe Village
To simulate the effect of rainfall pattern on dry matter and on grain accumulation
using the PT Model.
Methodology
The main method employed in studying the problem was yield simulation using PT
model. This method was supplemented with field visit by agricultural extension workers
and PT Help office staff. The field visit assisted in understanding the problems faced by
farmers in the village.
Field visit
Before the visit, the agricultural extension officers organized a village meeting with
Fulwe village farmers. The meeting was attended by village leaders and village extension
workers in addition to agricultural extension officers from the Morogoro and Mvomero
Districts and staff from the PT Help office. Figure 2 below shows some of the farmers at
Fulwe village in addition to researchers from PT Help office – Sokoine University of
Agriculture, agricultural extension officers from Morogoro and Mvomero Districts and
village extension worker outside Fulwe village office.
Figure 2. Fulwe village farmers
PT Model Help Office 17
Sokoine University of Agriculture
Case Study Two
Morogoro District – Fulwe Village
During the field visit, farmers were given opportunities to express the problems they are
facing with regard to maize production (Figure 3). The farmers reported that, agriculture
activities at Fulwe are seriously affected by poor rainfall, poor mechanization techniques,
lack of certified maize seeds and soil salinity.
Figure 3: One of Fulwe village farmers explaining problems facing their Agriculturalactivities
Input Data Collection
Weather data were obtained from Kingolwira Prison weather station and covered the
period from 1999 to 2004. The parameters collected were daily rainfall, maximum and
minimum temperature, humidity, radiation, wind speed and evaporation. Before running
simulations the data were formatted to the PT input standard. Soil profile data used was
also taken from Kingolwira (because of similarity to the soils found at Fulwe and
closeness to Fulwe village) and maize variety (TMV1) data was used.
Data Analysis
Simulations for different planting dates were run and summary and daily output files
were saved in the PT output folder. Output parameters analyzed were daily, seasonal and
cumulative rainfall, maize yield, soil moisture, dry matter accumulation and grain weight
gain.
PT Model Help Office 18
Sokoine University of Agriculture
Case Study Two
Morogoro District – Fulwe Village
Results and Discussion
Rainfall Patterns
The important growing window for maize in Fulwe is between January and May. Figure
4 shows daily rainfall trend over the long rainy season for different years.
1999 2000
2002 2003
0
10
20
30
40
50
60
70
80
90
1-Ja
n-00
21-J
an-0
0
10-F
eb-0
0
1-M
ar-0
0
21-M
ar-0
0
10-A
pr-0
0
30-A
pr-0
0
20-M
ay-0
0
9-Ju
n-00
29-J
un-0
0
Dates
Daily
rain
fall(
mm
)
0
5
10
15
20
25
30
35
40
1-Ja
n-99
21-J
an-9
9
10-F
eb-9
9
2-M
ar-9
9
22-M
ar-9
9
11-A
pr-9
9
1-M
ay-9
9
21-M
ay-9
9
10-J
un-9
9
30-J
un-9
9
Date
Rai
nfal
l (m
m)
0
10
20
30
40
50
60
1-Ja
n-02
21-J
an-0
2
10-F
eb-0
2
2-M
ar-0
2
22-M
ar-0
2
11-A
pr-0
2
1-M
ay-0
2
21-M
ay-0
2
10-J
un-0
2
30-J
un-0
2
Dates
Dai
ly ra
infa
ll(m
m)
0
10
20
30
40
50
60
1-Ja
n-03
21-J
an-0
3
10-F
eb-0
3
2-M
ar-0
3
22-M
ar-0
3
11-A
pr-0
3
1-M
ay-0
3
21-M
ay-0
3
10-J
un-0
3
30-J
un-0
3
Dates
Dai
ly ra
infa
ll(m
m)
Figure 4: Daily rainfall patterns for the years 1999, 2000, 2002, and 2003.
These charts show a consistent pattern of dynamic two dry-spells between January and
first week of March. Dynamic here means the actual dates of occurrence of the dry spells
are not known. Farmers in Fulwe normally plant their maize between mid January and
mid February. The dynamic dry-spells also means the start of the season is highly
variable. It is therefore possible that some of the maize face moisture stress which affect
the final yield. Furthermore, the area gets good rains between mid March and April.
PT Model Help Office 19
Sokoine University of Agriculture
Case Study Two
Morogoro District – Fulwe Village
There is poor rainfall in May in terms of low rainfall and frequent dry spells. These
characteristics have a significant effect on the final yield because most of the grain
weight accumulation occurs in May.
Effects of planting dates on maize grain yields at Fulwe village
Maize yield was simulated for different planting dates from 1999 to 2004. The results are
summarized in Figure 6. The best planting dates that resulted with higher yields were
between 20 January and 12 February. The average yields were between 0.45 and 0.49
t/ha and maximum yields ranged between 0.61 and 0.78 t/ha. However, the maximum
yields were not occurring only on dates with maximum yield (around February 5) instead
they occurred anywhere between 20 January and 12 February (a three-week window).
With regard to simulated yields, the simulated maize yields agreed with those reported by
farmers, which averaged between 0.5 and 0.7 t/ha. Therefore it was observed that
planting dates might not be the primary or the only reason for the low yields. It was
therefore necessary to carry out an analysis of dry matter accumulation.
0.00
0.10
0.20
0.30
0.40
0.50
0.60
1-Ja
n-05
8-Ja
n-05
15-J
an-0
5
22-J
an-0
5
29-J
an-0
5
5-Fe
b-05
12-F
eb-0
5
19-F
eb-0
5
26-F
eb-0
5
5-M
ar-0
5
12-M
ar-0
5
Dates
Ave
rage
Yie
ld(t/
ha)
Figure 6: Average yields for different planting dates
PT Model Help Office 20
Sokoine University of Agriculture
Case Study Two
Morogoro District – Fulwe Village
Dry matter accumulation
Results of dry matter accumulation are shown in Figure 7 and Figure 8. The results
indicate that dry matter accumulations at different planting dates were not significantly
different for both 1999 and 2000. The same was observed on the other two years.
Variation in dry matter accumulation from one year to another, which is normal, was
observed in years 1999 and 2000. This was attributed mainly to the little amount of
rainfall received in the year 1999. However, dry matter accumulation for 1999 was lower
than that for year 2000 due to the following reasons:
The year 1999 had both little annual and seasonal rainfall compared to year 2000. For
example, the seasonal rainfall, as computed using the PT model, for 1999 was 441mm
versus 514mm in 2000, while total annual rainfall was 761mm for 1999 versus
1011mm for 2000.
The year 1999 had longer dry spells during the growing period as compared to that of
2000. For example, the year 1999-experienced dry spells of 16 days from February 18
to 5 March; while for the year 2000 there were only 5 days of dry spells within the
same period.
0
1
2
3
4
5
6
2-Ja
n-99
12-J
an-9
9
22-J
an-9
9
1-Fe
b-99
11-F
eb-9
9
21-F
eb-9
9
3-M
ar-9
9
13-M
ar-9
9
23-M
ar-9
9
2-Apr
-99
12-A
pr-9
9
22-A
pr-9
9
2-M
ay-9
9
12-M
ay-9
9
Date
Dry
Mat
ter A
ccum
ulat
ion
(t/ha
)
1-Feb 6-Feb 11-Feb
Figure 7: Dry matter accumulation in 1999
PT Model Help Office 21
Sokoine University of Agriculture
Case Study Two
Morogoro District – Fulwe Village
Since our interest was to assess if the different planting dates could affect the dry matter
accumulation, the difference was not apparent or clear as shown in Figures 7 and 8. The
science of crop development shows that a crop such as maize is not affected significantly
by water stress, during its early days of growth. The most critical stage is at flowering
and grain filling stages. This called for further investigations into the grain filling stage.
0
1
2
3
4
5
6
7
8
2-Ja
n-20
00
9-Ja
n-20
00
16-J
an-2
000
23-J
an-2
000
30-J
an-2
000
6-Fe
b-20
00
13-F
eb-2
000
20-F
eb-2
000
27-F
eb-2
000
5-M
ar-2
000
12-M
ar-2
000
19-M
ar-2
000
26-M
ar-2
000
2-Apr
-200
0
9-Apr
-200
0
16-A
pr-2
000
23-A
pr-2
000
30-A
pr-2
000
7-M
ay-2
000
Date
Dry
Mat
ter A
ccum
ulat
ion
(t/ha
)
26-Jan 1-Feb 6-Feb
Figure 8: Dry matter accumulation 2000
Grain weight accumulation
For maize, and many other crops, grain filling stage is one of the critical periods for water
requirements. Therefore availability of water at this period would increase partitioning of
dry matter to grain leading to good crop yield. Since dry matter accumulation could not
show the significant difference observed in the maize yield, analysis of grain weight
accumulation was necessary.
Figure 9 and 10 show four maize grain filling charts for years 1999, 2000, 2002 and 2003
on a daily time step. Each chart has three lines representing three different planting dates.
It is clear from the charts that each date responded differently. All charts have zero values
up to late March or early April when the lines start to increase (grain filling start). The
earlier the start of grain filling the earlier was the planting/sowing date. Therefore, the
start of grain filling is fixed for a particular variety of maize. However, the start of the
PT Model Help Office 22
Sokoine University of Agriculture
Case Study Two
Morogoro District – Fulwe Village
grain filling stage and subsequent days are very important to the final yield. For example,
the comparison of grain filling rate for maize planted on 1st February and those planted
on 6th February show that at the first few days of grain accumulation the maize planted on
1st February was being affected by very short dry spells compared to maize planted on 6th
February. The grain filling was severely affected from 10th February because of a much
longer dry-spell. The same dry-spell period showed that the rate of filling affected most
the older plants than the younger ones since the growth rate of the maize planted on the
11th February was higher than that planted either on 1st or 6th February.
The second chart (year 2000) in Figure 9 shows a steady increase in grain weight
compared to the 1999 chart, where the increase is step-wise. The steeper slopes for maize
planted on 26th January and 6th February indicates a good supply of water during the start
of grain filling. There was a dry-spell during the start of the grain filling stage for maize
planted on the 1st February.
The 2002 chart in Figure 10 shows that all the lines are steadily increasing without any
cross-over and the first line (21-Jan) having a steeper slope and therefore higher yield
than the other two lines. Under normal circumstances, when the rainfall distribution is
good even during grain filling stage, this is supposed to happen. The reason for the lower
yields for the maize crop planted on 28th January and 1st February is the decrease in the
rainfall amount toward the end of the rainfall season. This is contrary to the 2003 chart in
Figure 10 because there is a cross-over between the 1st February and 11th February lines.
The maize planted on the 1st February (1-Feb) suffered severe water stress at the start of
grain filling stage. Maize planted on the 11th February (11-Feb) did not suffer water stress
because the rainfall came on time such that the rate of grain filling was very high. The
intersection of the grain filling lines was caused by the fact that the younger plants tend to
have the highest growth rate compared to the older ones under the same and good
environmental conditions (no water stress in this case). Therefore, when the water stress
came, already the 11-Feb maize had achieved good yield and the 1-Feb grain
accumulation had already been limited because of the previous dry-spell.
PT Model Help Office 23
Sokoine University of Agriculture
Case Study Two
Morogoro District – Fulwe Village
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
2-Ja
n-00
9-Ja
n-00
16-J
an-0
0
23-J
an-0
0
30-J
an-0
0
6-Fe
b-00
13-F
eb-0
0
20-F
eb-0
0
27-F
eb-0
0
5-M
ar-0
0
12-M
ar-0
0
19-M
ar-0
0
26-M
ar-0
0
2-A
pr-0
0
9-A
pr-0
0
16-A
pr-0
0
23-A
pr-0
0
30-A
pr-0
0
7-M
ay-0
0
Date
Gra
in W
eigh
t (t/h
a)
26-Jan 1-Feb 6-Feb
0
0.1
0.2
0.3
0.4
0.5
0.6
2-Ja
n-99
12-J
an-9
9
22-J
an-9
9
1-Fe
b-99
11-F
eb-9
9
21-F
eb-9
9
3-M
ar-9
9
13-M
ar-9
9
23-M
ar-9
9
2-A
pr-9
9
12-A
pr-9
9
22-A
pr-9
9
2-M
ay-9
9
12-M
ay-9
9
Date
Gra
in W
eigh
t (t/h
a)
6-Feb 1-Feb 11-Feb
Figure 9: Maize grain filling for 1999 & 2000 season
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
2-Ja
n-02
12-J
an-0
2
22-J
an-0
2
1-Fe
b-02
11-F
eb-0
2
21-F
eb-0
2
3-M
ar-0
2
13-M
ar-0
2
23-M
ar-0
2
2-A
pr-0
2
12-A
pr-0
2
22-A
pr-0
2
2-M
ay-0
2
date
Gra
in w
eigh
t(t/h
a)
21-Jan 26-Jan 1-Feb
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
2-Ja
n-03
12-J
an-0
3
22-J
an-0
3
1-Fe
b-03
11-F
eb-0
3
21-F
eb-0
3
3-M
ar-0
3
13-M
ar-0
3
23-M
ar-0
3
2-A
pr-0
3
12-A
pr-0
3
22-A
pr-0
3
2-M
ay-0
3
12-M
ay-0
3
22-M
ay-0
3
1-Ju
n-03
Date
Gra
in w
eigh
t (t/h
a)
1-Feb 11-Feb 21-Feb
Figure10 Maize Grain filling for 2002 and 2003 seasons
Therefore, the low and variable maize yields that farmers were experiencing in the village
were due to water stress during grain filling stage. Late planting also can result into lower
yields if a normal rainfall pattern occurs, which means decrease in the amount of rainfall
towards the end of the season.
Conclusions and Recommendations
Conclusions
Given soil characteristics and weather data of Kingolwira prison weather station, it
was possible to use the PT Model to determine appropriate management practices that
could increase maize yields. The practices include among others, conservation
agriculture and rainwater harvestingPT Model Help Office 24
Sokoine University of Agriculture
Case Study Two
Morogoro District – Fulwe Village
PT Model Help Office 25
Sokoine University of Agriculture
Dry matter accumulations for maize planted on different planting dates was not
significantly different for any particular year.
It was observed that grain weight (t/ha) differed for maize planted on different
planting dates. The main reason leading to the differences were due to dry-spells
during the grain filling stage.
Recommendations
Farmers at Fulwe village should practice conservation agriculture that will assist to
conserve soil moisture.
Also farmers should practice rainwater-harvesting (RWH) with storage techniques.
The stored water can later be used at the grain filling stage, which is the critical stage
of moisture requirements for maize crop.
Further testing/ using the PT Model to investigate the cause of poor maize crop yields
in other locations.
References
http://www.mapsofworld.com/lat_long/tanzania-lat-long.html
Mzirai, O.B.,T. Bwana, S.D. Tumbo, F.B. Rwehumbiza and H.F. Mahoo (2004), PARCHED THIRST Model Handbook, SWMRG – PT Help Office.
Robin M., S. William (2002), Crop-Soil Simulation Models Applications in Developing Countries, CABI Publishing, UK