1327 - ftcs and farmer tef demonstration and results 2012
TRANSCRIPT
2012 FTCs and farmers tef technology demonstrationOverview of results
May 2013
Agenda
A. Overview of results at federal levelB. Breakdown of results by regionC. Supplemental analyses
2
3
Newly-scaled up tef technologies, particularly row planting and transplanting, demonstrate yield improvements, with an average yield increase of ~70% over the national average
Average yield by planting method Quintals/hectare • Data was collected from
~15,800 validating farmers (and some control farmer groups) to determine the results of new tef technologies
• For the 14,605 farmers, average yields for row planting and transplanting increased 70% from national average (20.9 versus 12.6 qtls/ha)
• As the chart shows, there is still much work to be done in properly managing transplanting to realize potential yield gains
17
12
16
21222122
1820
2323
Amharan = 4,637
SNNPn = 3,480
Oromian = 6,002
Tigrayn = 486
N/A
BroadcastingRow plantingTransplanting
Source: 2012 Data from Regional, Zonal and Woreda administration staff (collected Feb-April 2013)Note: Includes data from 14,605 farmers (omitted error/outlier data from 15,790 total collected)
4
Distribution of yield data shows that 30% of all validating farmers surveyed experienced yield increase between 20 and 80% over the national average
Source: 2012 Data from Regional, Zonal and Woreda administration staff (collected Feb-April 2013)Note: Includes data from 14,605 farmers (omitted error/outlier data from 15,790 total collected)
Distribution of Validating Farmers’ yieldsFrequency of yield increase (as % of total data set)
476
810
8
1311
7
150 - 200%
125 - 150%
Over 200%
100 - 125%
80 - 100%
60 - 80%40 - 60%20 - 40% 10 - 20%Less than 10%
27~30% of farmers saw a 20 – 80% yield increase
Farmers who broadcasted, used high seed rates, or may have experienced challenges
with new technologies
~20% of farmers saw a 100 – 200% yield increase (~60% of this
group row planted)
5
Top-performing woredas across the four regions have demonstrated success in adopting the new technologies, reaching maximum yields higher than a 400% yield increase
Average and maximum yields for row planting and transplanting farmersQuintal/hectare
36
36
35
35
35
35
33
33
32
31H/Abote 53
Tahaty-Maichew 56
47
Lume 68
Maraka
Kuxha
68
65
Alefa 42
Dabat
EnebsieSar Midir 58
Chilga 53
67
Libo KemkemAMHARA
SNNP
OROMIA
SNNP
TIGRAY
OROMIA
68 quintals/hectare versus the national
average of 12.5 quintals/hectare
Source: 2012 Data from Regional, Zonal and Woreda administration staff (collected Feb-April 2013)Note: Includes data from 14,605 farmers (omitted error/outlier data from 15,790 total collected)
6
Data indicates that validating farmers’ seed rate usage varies by planting method, with the highest rates used in broadcasting to the least amount of seed used in transplanting
Average seed rate and yields by planting type across regionsYield (quintal/hectare), Seed rate (kg/hectare)
4
1011
302222
1513
25
20
15
10
5
0
30
25
20
15
10
5
0TransplantingRow plantingBroadcastingNational
average
• While this does not imply causation, there appears to be an inverse relationship between planting type and seed rate practices of the validating farmers
• Overall, the average seed rate has dropped significantly from traditional practices of 30-50 kg/ha to an average of 8.8 kg/ha across all planting types (includes broadcasting farmers as well)
• Use of Quncho variety appears to be on the rise, with 92% of validating farmers choosing it over other options (local varieties, Cross-37, Yedega, etc.)
Seed rateAvg yield
Source: 2012 Data from Regional, Zonal and Woreda administration staff (collected Feb-April 2013)Note: Includes data from 14,605 farmers (omitted error/outlier data from 15,790 total collected)
7
Fertilizer application rates appear to match recommendations for DAP (average application rate of 99.5 kg/ha), yet fall slightly below recommendations on Urea (88.4 kg/ha)
Average DAP and Urea application rates by regionKg/hectare
Key takeaways on DAP and Urea utilization by farmers: • Urea use is highly variable,
much more than DAP: the standard deviation for Urea is 24.6 kg/hectare versus 3.8 for DAP, indicating the that farmer use of DAP is more consistent
• Application rates are fairly agnostic of planting method: for each planting method, the average application rates remain similar
991009991868890
Amharan = 4,037
SNNPn = 3,051
Oromian = 4,943
Tigrayn = 304
N/A
DAPUrea
4.1 21.9 4.4 25.4 1.3 26.5 6.5 21.5
X Standard deviation
Source: 2012 Data from Regional, Zonal and Woreda administration staff (collected Feb-April 2013)Note: Includes data from 12,335 farmers (omitted error/outlier fertilizer data from 15,790 total collected)
8
A high-level view of planting time indicates that most farmers observed ‘traditional’ planting times, though farmers that planted earlier saw significant yield increases
6961626162627286
668089
0
50
100
June Week 2
SeptWeek 1
July Week 4
AugustWeek 1
August Week 2
August Week 3
AugustWeek 4
59
JulyWeek 3
July Week 2
July Week 1
JuneWeek 4
June Week 3
Average productivity increase by planting time% productivity increase (compared to national average)
0.3%0.1% 2%0.8% 16%5% 19%24% 4%7% 1%4%
Perc
ent o
f fa
rmer
s
Key takeaways Planting time of validating farmers followed
standard practices, with ~60% of farmers planting in mid-July to early August
Though a small sample size, there seems to be a strong productivity increase for farmers who planted earlier, in the 3 and 4th weeks of June
Standard planting period
Next steps for exploration Early planting time, when is coincides with
rainfall, is a topic to be further explored with farmers and on FTC plots
EIAR and RARIs, with support from ATA, will conduct formal research projects to determine the impact of planting 2, 3 or 4 weeks early
Source: 2012 Data from Regional, Zonal and Woreda administration staff (collected Feb-April 2013)Note: Includes data from 12,335 farmers (omitted error/outlier data from 15,790 total collected)
Largely driven by transplanting
9
20.118.3
15.5
8.4
12.6
+60%
Average yields by experimental plot across regionsQuintal/hectare
X.X Standard deviation
Avg yield qt/ha
Planting type
National Average for
2012
Broadcast by hand
Broadcast by machine
Row plant Transplant
Seed rate (kg/ha) 30-50 5-10 5-10 0.5-0.7
Fertilizer type none DAP +Urea
DAP + Urea
DAP +Urea
Seed type Local Quncho Quncho Quncho
8.16.9 9.4 11.4
In ~1,100 FTC trials, it can be observed that factors such as reduced seed rate, planting in rows, and use of DAP and Urea can each contribute to increasing yields
• Findings indicate that significant yield
improvement results from new technologies such as row planting,
transplanting, and reduced seed rate
• Across the 1,100 FTCs, yields were recorded for each experimental plot
that was designed to test either planting
method, seed rate, seed variety, or fertilizer use
Source: 2012 Data from Regional, Zonal and Woreda administration staff (collected Feb-April 2013); CSA 2012
10
20.118.3
15.514.913.912.012.6
+4%+7%
+18%+10%
+16%
Average yields by experimental plot across regionsQuintal/hectare
Planting type
National Average for
2012
Broadcast by hand
Broadcast by hand
Broadcast by hand
Broadcast by machine
Row plant Transplant
Seed rate (kg/ha)
30-50 30-50 5-10 5-10 5-10 0.5-0.7
Fertilizer type
DAP + Urea DAP +Urea
DAP + Urea
DAP +Urea
DAP +Urea
DAP +Urea
Seed type Local Quncho Quncho Quncho Quncho Quncho
Taking a closer look at each variable, it appears that switching to Quncho and row planting are the largest drivers of productivity increase
Source: 2012 Data from Regional, Zonal and Woreda administration staff (collected Feb-April 2013); CSA 2012
Agenda
A. Overview of results at federal levelB. Breakdown of results by regionC. Supplemental analyses
11
12
AMHARA: overview of data collected
Source: 2012 Data from Regional, Zonal and Woreda administration staff (collected Feb-April 2013)
Farmer data overview
Average yield 20.47
Standard deviation 9.65
Average productivity increase 62%
Count of farmers Raw Cleaned
Farmers 4,958 4,637
Gender split
Male 4,373 95%
Female 685 4%
Geographic scope within region (as reported)
Count of zones 10
Count of woredas 85
31%
Broadcasting (machine)1%Broadcasting
(hand)
Transplanting
Row planting48%
19%
Distribution of planting type% of total farmers (from raw data)
13
AMHARA: input use, yields achieved, and planting time
Source: 2012 Data from Regional, Zonal and Woreda administration staff (collected Feb-April 2013)
Inputs use of Validating Farmers
Planting time of Validating Farmers
Highest-producing woredas (based on Validating Farmers)
Average input use by planting method Kg/hectare
10
889994
9
99 96100
2
Row plantingn = 2,238
Broadcastingn = 1,446
Transplantingn = 889
Seed
DAP
UreaAverage and maximum farmer yieldsKg/hectare (average and max by woreda)
36
35
32
31
30
Enebsie Sar midir 58
Alefa
Dabat 67
Libo Kemkem 68
53Chilga
42
AverageMax
Average yield based on planting time% productivity increase (compared to national average)
68194136455671947285105104
100
August Week 3
200
0Sept
Week 1AugustWeek 4
August Week 2
AugustWeek 1
July Week 4
JulyWeek 3
July Week 2
July Week 1
JuneWeek 4
June Week 3
June Week 2
14
18.016.3
14.5
8.0
12.6
+124%
Average yields by experimental plots across Amhara FTCsQuintal/hectare
X.X Standard deviation
Avg yield qt/ha
Planting type
National Average for
2012
Broadcast by hand
Broadcast by machine
Row plant Transplant
Seed rate (kg/ha) 30-40 5-10 5-10 0.5-0.7
Fertilizer type none DAP +Urea
DAP + Urea
DAP +Urea
Seed type Local Quncho Quncho Quncho
8.15.0 8.9 10.4
AMHARA: FTC demonstration plot results
Source: 2012 Data from Regional, Zonal and Woreda administration staff (collected Feb-April 2013); CSA 2012
15
OROMIA: overview of data collected
Source: 2012 Data from Regional, Zonal and Woreda administration staff (collected Feb-April 2013)
Farmer data overview
Average yield 21.26
Standard deviation 8.56
Average productivity increase 69%
Count of farmers and FTCs Raw Cleaned
Farmers 6,515 6,002
Gender split
Male 6,198 95%
Female 292 5%
Geographic scope within region (as reported)
Count of zones N/A
Count of woredas 48
15%Broadcasting
57%Transplanting
Row planting
16%
Distribution of planting type% of total farmers (from raw data)
16
OROMIA: input use, yields achieved, and planting time
Source: 2012 Data from Regional, Zonal and Woreda administration staff (collected Feb-April 2013)
Inputs use of Validating Farmers
Planting time of Validating Farmers
Highest-producing woredas (based on Validating Farmers)
Average input use by planting method Kg/hectare
4810
1009999 859085
Transplantingn = 889
Row plantingn = 2,238
Broadcastingn = 1,446
Urea
DAP
SeedAverage and maximum farmer yieldsKg/hectare (average and max by woreda)
32
30
30
28
26Dawoo 38
A/Nagale 65
Gimbichuu 61
H/Abote 53
Lume 68
AverageMax
Average yield based on planting time% productivity increase (compared to national average)
9262616467677290
667551
2050
August Week 3
August Week 2
AugustWeek 1
July Week 4
JulyWeek 3
July Week 2
100
0Sept
Week 1AugustWeek 4
July Week 1
JuneWeek 4
June Week 3
June Week 2
17
20.118.1
15.2
8.1
12.6
+148%
Average yields by experimental plots across Oromia FTCsQuintal/hectare
X.X Standard deviation
Avg yield qt/ha
Planting type
National Average for
2012
Broadcast by hand
Broadcast by machine
Row plant Transplant
Seed rate (kg/ha) 30-40 5-10 5-10 0.5-0.7
Fertilizer type none DAP +Urea
DAP + Urea
DAP +Urea
Seed type Local Quncho Quncho Quncho
8.46.4 9.5 11.1
OROMIA: FTC demonstration plot results
Source: 2012 Data from Regional, Zonal and Woreda administration staff (collected Feb-April 2013); CSA 2012
18
SNNP: overview of data collected
Source: 2012 Data from Regional, Zonal and Woreda administration staff (collected Feb-April 2013)
Farmer data overview
Average yield 20.72
Standard deviation 9.06
Average productivity increase 64%
Count of farmers and FTCs Raw Cleaned
Farmers 3,766 3,480
Gender split
Male 3,061 81%
Female 124 3%
Geographic scope within region (as reported)
Count of zones N/A
Count of woredas 38
11%
65%Row planting
1%
Broadcasting(hand) Broadcasting (machine)
13%Transplanting
Distribution of planting type% of total farmers (from raw data)
19
SNNP: input use, yields achieved, and planting time
Source: 2012 Data from Regional, Zonal and Woreda administration staff (collected Feb-April 2013)
Inputs use of Validating Farmers
Planting time of Validating Farmers
Highest-producing woredas (based on Validating Farmers)
Average input use by planting method Kg/hectare
41211
100100100838783
Transplantingn = 889
Row plantingn = 2,238
Broadcastingn = 1,446
Urea
DAP
SeedAverage and maximum farmer yieldsKg/hectare (average and max by woreda)
33
33
27
27
27Bona Zuria 33
Maraka 47
Gombera 30
Kuxha 70
Dalocha 69
AverageMax
Average yield based on planting time% productivity increase (compared to national average)
686367666972725234178
-37144
-200
200
0
SeptWeek 1
AugustWeek 4
August Week 3
August Week 2
AugustWeek 1
July Week 4
JulyWeek 3
July Week 2
July Week 1
JuneWeek 4
June Week 3
June Week 2
June planting consists of only 5 farmers
20
22.721.0
17.0
8.2
12.6
+177%
Average yields by experimental plots across SNNP FTCsQuintal/hectare
X.X Standard deviation
Avg yield qt/ha
Planting type
National Average for
2012
Broadcast by hand
Broadcast by machine
Row plant Transplant
Seed rate (kg/ha) 30-40 5-10 5-10 0.5-0.7
Fertilizer type none DAP +Urea
DAP + Urea
DAP +Urea
Seed type Local Quncho Quncho Quncho
7.27.8 8.9 12.5
SNNP: FTC demonstration plot results
Source: 2012 Data from Regional, Zonal and Woreda administration staff (collected Feb-April 2013); CSA 2012
21
TIGRAY: overview of data collected
Source: 2012 Data from Regional, Zonal and Woreda administration staff (collected Feb-April 2013)
Farmer data overview
Average yield 20.55
Standard deviation 9.09
Average productivity increase 63%
Count of farmers and FTCs Raw Cleaned
Farmers 551 486
Gender split
Male 410 74%
Female 54 10%
Geographic scope within region (as reported)
Count of zones N/A
Count of woredas 11
4%
62%
Row planting
5%Transplanting 4%
Broadcasting(hand) Broadcasting (machine)
Distribution of planting type% of total farmers (from raw data)
22
TIGRAY: input use, yields achieved, and planting time
Source: 2012 Data from Regional, Zonal and Woreda administration staff (collected Feb-April 2013)
Inputs use of Validating Farmers
Planting time of Validating Farmers
Highest-producing woredas (based on Validating Farmers)
Average input use by planting method Kg/hectare
7813
10098100 9192100
Transplantingn = 889
Row plantingn = 2,238
Broadcastingn = 1,446
Urea
DAP
SeedAverage and maximum farmer yieldsKg/hectare (average and max by woreda)
32
30
30
28
26Tahtay Koraro 38
Weri-Leke 65
Medebay Zana 61
Adwa 53
Tahaty- Maichew 68
AverageMax
Average yield based on planting time% productivity increase (compared to national average)
367449
-2529100
0
SeptWeek 1
AugustWeek 4
August Week 3
-100August Week 2
AugustWeek 1
July Week 4
JulyWeek 3
July Week 2
July Week 1
75
JuneWeek 4
-17
June Week 3
67
June Week 2
7636N/A N/A
Includes only 2 farmers
Includes only 2 farmers
23
18.918.716.5
11.812.6
+60%
Average yields by experimental plots across Tigray FTCsQuintal/hectare
X.X Standard deviation
Avg yield qt/ha
Planting type
National Average for
2012
Broadcast by hand
Broadcast by machine
Row plant Transplant
Seed rate (kg/ha) 30-40 5-10 5-10 0.5-0.7
Fertilizer type none DAP +Urea
DAP + Urea
DAP +Urea
Seed type Local Quncho Quncho Quncho
9.511.0 10.5 11.8
TIGRAY: FTC demonstration plot results
Source: 2012 Data from Regional, Zonal and Woreda administration staff (collected Feb-April 2013); CSA 2012
Takeaways from 2012 farmer and FTC data analysis that should inform ATA and RBoA planning and execution for 2013 and 2013 tef planting seasons
24
Takeaways to inform 2013 and 2014 intervention planning and execution
• The benefits of optimal fertilizer application rates should be better disseminated to farmers via trainings and materials, particularly for Urea
• More intensive training on transplanting management practices should be provided to farmers given the volatility in yields observed in 2012
• FTCs should be encouraged to conduct demonstrations to test for specific variables’ impact on yield (e.g., planting 2-3 weeks earlier, reduced seed rates of 5/10/15/ kg per ha, pelleted Urea)
• FTC yields indicate significant yield increases due to use of Quncho with 90% of farmers in these trials using Quncho; should explore further support for Quncho while also considering other improved varieties
25
Includes 15,790 validating farmers’ data
8% of data was omitted as inaccurate or incomplete
45% (33% + 12%) of validating farmers’ yields were between 20 and 40 quintals per hectare
Overview of data collected: yield distribution for 15,790 validating farmers and 1,107 FTCs
Source: 2012 Data from Regional, Zonal and Woreda administration staff (collected Feb-April 2013)
40 to 60
40
0
20
1%
60+
3%
30 to 40
12%
20 to 30
33%
10 to 20
40%
Less than 10
8%
Distribution of Validating Farmers’ yieldsFrequency of yield increase (as % of total data set)
Covers 1,107 FTCs with an average of 9.6 experimental data plots per FTC
4% of the data was omitted as inaccurate or incomplete
Distribution of yields is wider given trials included many variables
40
20
040+
2%
30 to 40
7%
20 to 30
20%
10 to 20
38%
5 to 10
18%
Less than 5
11%
Distribution of FTC yield per experimental plotFrequency of yield increase (as % of total data set)
Quintals / hectare
Quintals / hectare
26
Overview of data collected: basics
Source: 2012 Data from Regional, Zonal and Woreda administration staff (collected Feb-April 2013)
Farmer data overview
average yield 20.86std dev 12.38bell curve upper bound 58.00
average productivity increase 66%
Count of farmersAmhara 4,958 31%Oromia 6,515 41%SNNP 3,766 24%Tigray 551 3%Total farmers 15,790
Gender splitMale 14346 95%Female 685 5%
20%1%
17% 61%
Broadcasting (machine)
Row plantingTransplanting
Broadcasting (hand)
Distribution of planting type% of total farmers