Amibara Drainage Project - Irrigation Agronomy Report Summary

The People’s Democratic Republic of Ethiopia
Water Resources Development Authority
J
V" ■
.X
IRRIGATION AGRONOMY
REPORT
JULY - AUGUST 1989 FINAL DRAFT
AMIBARA DRAINAGE PROJECT I
Sir M. MacDonald & Partners Limited
Cambridge, EnglandLIST OF CONfENTS
PAGE
1.         ITINERARY AND WORK PROGRAMME............................................................... 1
2.         RECENT PRODUCTION TRENDS .AND PROJECT MONITORING.................... 3
2.1       Introduction.....................................................................................................
2.2       Production Trends on the Melka Sadi                    State Farm... 
3
2.2.1.          Cotton yields..................................................................................................  3
2.2.2.          Banana yields................................................................................................  4
2.3       Climatic Data..................................................................................................................... 4
2.4       Groundwater Levels and Soil Salinity............................................................................... 3
2.5       Irrigation Water Supply..................................................................................................... 5
2.6       Irrigation Water Quality....................................................................................................  6
3.      IDENTIFICATION OF FACTORS LIMITING CROP PRODUCTION
LEVEIES..................................................................................................................................... 14
3.1       Introduction..................................................................................................................... 14
3.2      The Incidence of Soil Salinity......................................................................................... 14
3.3      The Causes of             Soil Salinity.................................................................................  15
3.4      Resource Li mi tations.....................................................................................................  18
3.4.1.          Water supply...............................................................................................  18
3.4.2.          Machinery and equipment..........................................................................   19
3.4.3.          Availability of
capital........................................................... 19
3.4.1.          Human resources.........................................................................................  19
3.5 Husbandry Li mi tations............................................. ..........................................  21
3.5.1.          Bananas................................ *.....................................................  21
3.5.2.          Cotton..........................................................................................  21
3.6       The Amibara Settlement Project.....................................................................  22
4.    FUTURE AGRICULTURAL DEVELOPMENT REQUIREMENT'S...........................  24
4.1       Objectives and Policies.................................................................................... 24
4.2       Land Levelling................................................................................................................ 26
4.2.1.          Background.................................................................................................  26
4.2.2.          Extent of land levelling requirements.... 
28
4.2.3.          Options for land 
levelling............................................................ 28
4.3       Land R*icl amat ion.......................................................................................................  28
4.4       Improvement of Water Management               Practices...............................................  29
4.5       Improvement of Land Preparation 
and Crop
Husbandry Practices.......................................................................................................  30
4.5.1.          Bananas.......................................................................................................  30
4.5.2.          Cotton.........................................................................................................  31
4.6       Agronomic Trial Programme.......................................................................................... 32
4.7       Training........................................................................................................................... 33
4.8       Amibara Settlement Project                   Planning..........................................................  34LIST OF CONTENTS (CONTD.)
5.        PROPOSALS AND RECOMMENDATIONS - ACTION PLAN FOR 1990.. 5.1       Project         Monitoring Improvements and Innovations.. 
PAGE
36
36
5.1.1.          Soil salinity monitoring...........................................................................  37
5.1.2.          Aerial photography..................................................................................  37
5.1.3.          Monitoring water quality.........................................................................  37
5.1.4.          Monitoring crop production levels............................................               37
5.2       Definition of Size and Scope of Land Levelling
Requirements..............................................................................................................  38
5.3       Water Management Observations and Measurements... 
38
5.4      Observation on Groundwater Fluctuation................................................................... 39
5.5       Applied     Research...................................................................................................  39
5.5.1.          Leaching and land             reclamation..................................................  39
5.5.2.          Bananas...................................................................................................  40
5.5.3.          Cotton.....................................................................................................  41
5.5.4.          Miscellaneous trials................................................................................  41
5.5.5.          Implementation of             a crop trial programme. 
G. PRELIMINARY REVIEW OF THE IRRIGATION AGRONOMY REPORT
BY OTHER ORGANISATION...............................................................................  43
42
6.1 Distribution of the Report and Objectives..................................................... ..                  43APPENDIX   NO. 1
2
LIST OF APPENDICES
Comments on report by Institute of Agricultural Research Comments on report by Project Control CentreLIST OF TABLES
TABLE NO. la
lb
2
3 3a 3b 3c
4
Mean     Yields     of     Seed     Cotton     at     the     Melka     Sadi     State     Fann     by Management Unit and Some Results of Statistical Analyses
Melka Sadi State Fann Cotton Production Banana Production at the Melka Sadi State Farm
Monthly Rainfall
Data from Stations in the ADP Stage 1 Area
5a
5b 6a 6b 7
IAR Agroclimatic            Station
PCC Agroclimatic            Station
Pilot Drainage Trial Rain Gauge
Number of Piezometers by Depth to Water Table Ranges
Illustrating Fluctuations in Groundwater Levels Between 
1984-89
Annual      Summary      of      Irrigation      Water      Supplied      to      the      Three
Users in the Amibara Irrigation Area
Mean         Annual         Quantities         of         Water         Delivered         Through         Primary Offtakes to Melka Sadi State Farm
River Water Salinity as Monthly Mean Electrical Conductivity
Actual      Measurements      of      River      Water      Salinity      as      EC      in 
niS per cm
Inventory      of      Farm      Machinery      at      the      Melka      Sadi      State      Fann, July 1989
*2ITINERARY AND WORK PROGRAMME
The Irrigation Agronomist left UK on July 10 arriving in Addis
Ababa the following day and travelling on to the Aniibara Project on 
July 14. He was on site until August 3 when he returned to Addis
Abuba and left for the UK on August 4.
The terms of reference for the Irrigation  Agronomist which  are set out in  the Technical Proposal which  forms the basis for the
Consul Laney Contract Agreement include the responsibilities:
following
a.                 Reviewing        present        ci’op        production        practices        to        identify        crop yield limiting factors.
b.               Assembling       crop       production       data,       related       to       soil       conditions       for establishing priority areas for investigation.
c.         Establishing a monitoring system for crop yield assessment in 
relation to soil conditions, irrigation water quality und water
management.
d.                Assisting       with       the       selection       and       planning       of       a       pilot       druin trial on vertisol soils.
e. 
Following  land  reclamation  preparing  cropping  pitterns, irrigation  scheduling,  land  management technique and  field  lay outs .
f.                 Planning         and         supervising         implementation         of         field         t.rials         to 
investigate            alternative            methods            of            surface
irrigation 
the
effects           of           variable           irrigation           scheduling           on           soil           salinity levels and crop yields.
In  addition  to  these terms of reference a list of*  priorities to  be covered  by  the Irrigation  Agronomist during  his several visits were provided  by  the Consultants Supervisor and  Team leuder which  were as follows:
investigate the scope for a second crop
identify         the         need         for         improved         water         management         particularly scheduling for various growth sluges.
level          ling/planing          -          equipment,          subsoiling,          compact          soil layers, inter-row recommendations.
leaching programmes
improved land preparation
crop husbandry/banana management
moniLoring system for crop yields
Both sets of terms of reference are wide ranging and whilst they overlap to a large extent they obviously could noL be comprehensively tackled in this first visit.
1The basic aims of this visit have been  to  obtain  an  understanding  of the situation  in  the Stage I area in  the broadest sense and  to identify  priority  areas for action.  Existing  information  and  data on  all aspects of crop  production  have been  examined.  Institutional and  organisational responsibilities relating  to  the project have been  established.  Meetings have been  held  with  managers in  MAADE, TAR and  RRC to  establish  a working  relationship  and  to  take advantage of the accumulated  knowledge on  factors limiting  crop production•levels.
This report includes a brief summary  of the findings of the initial visit of the Irrigation  Agronomist to  ADP in  July  1989.  The Liming of this visit was most opportune in  so  far as it was possible to  see clearly  the problems of unlevel fields and  the problems of soil salinity  which  were obviously  well defined  in  the cotton.  However, 
with  the crop  already  well established  there is only  limited 
opportunity  for investigational work  and  in  any  case the
organisational, 
logistical
and 
funding 
complications
rule
out
any 
immediate
action. 
Consequently 
the
proposals
contained 
in 
this
report refer to  the consolidation  of existing  monitoring  procedures for the 1989  cotton  crop  whilst for 1990  a much  wide range of activities could  be possible including  land  development estimates, field  investigations and  crop  trials and  the introduction  of more sophisticated systems of monitoring.2.                     RECENT PRODUCTION TRENDS AND PROJECT MONITORING
2.1
Introduction
In  order to  have a proper understanding  of the existing  situation  it is necessary  to  up  date the report entitled  ’Master Drainage Plan for the Melka Sadi and  Amibara Areas'.  This report was published  in July  1985  but much  of the data presented  covered  the period  up  to 
1983      whilst      some      of      the      production      projections      used      were      based      on assessments       of       the       immature       1984       crop.       Since       1984       a       further       period of     five     years     has     elapsed     in     which     a     considerable     amount     of     data     and experience         has         accumulated.         This         accumulated         information         has         rarely been        analysed        although        it        is        a        vital        ingredient        of        the        planning process.
Within  the Stage 1  area of ADP there are four divisions of the land which  are under four district management systems: the Melka Sadi State Farm (4205  hectare irrigable); the Melka Werer IAR (300  ha); the Amibara Settlement Project (2000  ha); the Melka Werer State Farm,  (500  ha).  Tn  examining  recent production  trends the emphasis has been  placed  inevitably  on  the Melka Sadi State Farm,  the whole of which  is within  the Stage 1  area.  Conditions on  the I AR are atypical of the state farms.  The Amibara Settlement Project does not keep  records of cropping  and  crop  production.  Parts of the Melka Sadi State Fann  are representative of the small area of the Melka Werer State Farm within  the Stage 1  area and  it was considered unnecessary to duplicate.
The Melka Said  State Farm is situated  at the head  of the Amibara irrigation  system and  includes areas which  have been  irrigated  since the late 1960's as well as areas which  were developed  as late as 1984.  This farm includes substantial areas of abandoned  land  and  is therefore a suitable selection for more detailed study.
2.2                   Production Trends on the Melka Sadi State Fann
2.2.1                Cotton yields
Yield  data has been  collected  at the individual field  level in  Melka Sadi since the beginning  of the project in  1984  and  this dutu  lias been  carefully  analysed.  'Hie mean  yields by  management or irrigation  unit are presented  in  Table la for the five year period 1984-88.  Tn  the first three years according  to  these statistics
ovemll mean  yield  decreased  by  20% and  this was followed  by  some recovery in  1987.  However in  1988  there was a drastic reduction  in 
production      levels      when      the      uvenige      yield      was      only      60%      of      the      1984 
yield.
Slxitistical analyses were undertaken  of the yield  data for cat'll of the seven  units.  Thu  differences lie tween  the annual me; ins and between  field  menus were significant in  all seven  units nt least al the 5% level with  l.he majority  of inter year differences significant, at.  the 0.1% level.  The coefficieiil.s of variation  of the yield  data were within  acceptable limits which  sup|A>rl.s the conclusion  that the basic data is reasonably rdiublc.
3The conclusion  that the field  yield  data is reliable may  be challenged  on  two  points.  The first is the farm data contained  in Table lb  which  shows a more regular but accelerating  decline in yield  and  the second  stems from an  examination  of individual field data.  In  the latter,  fields which  are known  to  be seriously affected  by  soil salinity  do  not give appropriate yield  reductions or the pattern  of yield  reduction  is not one of gradual reduction but a catastrophic fall in  production.  However the number of fields which  have been  obviously  wrongly  recorded  is relatively  small arid do  not seriously  affect overall conclusions.  A catastrophic fall in a field  production  level may  be attributable to  factors other than soil salinity.
There is an  obvious trend  of declining  yields of cotton  on  the Nelka Sadi State Fann  based  on  an  examination  of the last 5  years of data. Field  observations suggest that this trend  is due large1y  to increasing  soil salinity  and  unless some action  is taken  to  reduce and control soil salinity levels yields will continue to decline.
2.2.2 Banana yields
The banana area at Melka Sadi State Fann  was develop'd  in  the late 1960’s and  the available production  figures which  are for- the period 1975-1988  are contained  in  Table 2.  These statistics demonstrate that the management has been  in  a losing  battle to  hall,  delining production  levels and  land  abandonment and  the survival of the whole
banana operation must now V productive area of bananas
be in danger. has fallen well
Indeed  in 
1989 the
below the  -110          hectares
that was maintained for the previous four years.
The basic problem with the
- water table which has led 
banana area has
to heavy salinisation of
been one  of
the
a rising
topsoil.
Other          major          constraints          have          included          lack          of          mint          ion, uncontrolled weed growth, and inadequate fertilisation.
2.3                   Climatic Data
Climatic data has been  examined  for the 5  year period  and exceptional events were largely  confined  to  rainfall.  Rainfall data is presented  in  Table 3  for- 3  stations.  The main  feature of this data is the extra ordinary  difference lie tween  the three stations and particularly  the much  higher rainfall recorded  at 1AR.  IVssililc explanlions for this variability  is given  in  App-ndix  2  of this repnt.
—2.4                   Groundwater Levels and Soil Salinity
Data from 71  piezometers that were installed  in  1984  have been examined.  These piezometers are located  in  a random manner but give good  coverage of both  the Melka Sadi and  Melka Merer State Farms. A broad  indication  of general changes in  groundwater depth  is illustrated  by  the data shown  in  Table 4.  In  this table a gradual trend  towards higher water table levels can  be detected  but there have been  no  massive permanent changes as were once predicated. Indeed  there are no  piezometers showing  consistently  high  water tables of less than  one metre.  It should  however be pointed  out that January  and  June may  not register the highest water table levels recorded through the year.
The piezometer data also  demonstrates that water table levels are fluctuating  often  by  1  or 2  metres and  occasionally  more,  thus suggesting that some natural drainage is taking place.
J
The largest rise in  water table levels occurred  in  June 1989  when 49.2% of the piezometers registered  water table levels of less than 3.0  m compared  with  36.9% in  June 1988.  There is no  ready explanation for this most recent change.
High           soi1 groundwater
salinity          levels          are          commonly levels and this is simply
associa   ted   w   i  t  h demonstrated by
abandoned saline areas of the banana area on the Melka Sadi
h i gh the
State
Fann.  Unfortunately  no  routine monitoring  of soil salinity  levels has been  undertaken  so  that there is no  definite proof of the correlation  between  groundwater level and  soil salinity  when*  the water table is more than 1 m below the surface and the extent of
transmission             of wuter to the surface by capillary uncertain.
act ion is
2.5 
Il must be acknowledged  however that the events (irrigation frequency) field  reinstatement etc.) under which  data is collected vury  from area to  area and  general conclusions art*  difficult to arr ive al.  A comprehensive analysis of groundwater depth  in conjunction  with  lurid  use details might derive boiler conclusions - (see Appendix 2).
Irrigation Water Supply
The data on  the quantities of waler passed  down  the main  canal and allocated  to  the primary  offtakes have been  ex;utdiied  with  the main objectives of giving  a broad  indication  of year h> year fluctuation in  water use and  Io  obtain  some estimates of the actual quantities of waler* made available on a jxt hectare taisis.
5In  Table 5a an  annual summary  is presented  of the quantities of water made available to  the three main  users - the Middle Awash Agricultural Enterprise,  the Research  Station  and  the Amibara Settlement Farm.  Annual water use shows little correlation  with rainfall in  the irrigated  area.  However the low figure for 1988  was due to  low river flows caused  by  low rainfall in  the River Awash catchment area in  the Central Highlands - dependence on  run  of the river water supply  may  explain  the absence of any  correlation between  irrigation  water use and  rainfall.  The low 1988  irrigation water quantity  is also  because a considerable area remained unplanted  because of the general water shortage during  pre irrigation and planting periods in that season.
Tn  Table 5b  the mean  annual quantities of water delivered  through the primary  offtakes serving  the Melka Sadi State Farm are presented expressed  on  a per hectare basis.  .  Rainfall data is also  included in  Table 5b  so  that the total amounts of water available for the cotton  can  be calculated.  No  firm conclusions can  be drawn  from
this      data      but      it      is      suggested      that,      with      a      net      requirement      of      8200 
Cu.  m per ha for cotton,  in  many  instances there is the tendency for over irrigation.  However irrigation  quantity  fluctuations from year to  year could  be attributed  to  the policy  of State Fanns to  try to introduce double cropping. (See Append* 2).
2.6 J Irrigation Water Quality
The routine recording  of Awash  river water quality  began  in  June 1987.  The electrical conductivity  of water samples taken  from the
Amibara     main     canal     headworks     are     being     measured     four     times     a     month at more or less weekly intervals.
In  the period  June to  December 1987  there was little change in  the electrical conductivity  of the water which  varied  between  0.34- 0.40  mS per cm.  However from the beginning  of 1988  a gradual increase occurred  which  continued  to  a peak  monthly  mean  of 0.88  mS per cm in  June and  an  overall peak  of 1.04  mS per cm in  the first week  in  July.  These high  figures coincided  with  exceptionally prolonged low river flows in 1988.
According  to  the USDA classification  of irrigation  water salinity the Awash  river water may  be generally  classified  us C2  medium salinity  water which  can  only  be used  on  a long  term basis if a moderate amount of leaching  occurs.  For short periods waler* quality  may  improve to  an  ECe of less than  0.25  mS per cm.  However in  1988  there was a period  when  water quality  moved  info  the C3 
category  described  as high  salinity  water in  which  ECe exceeded 0.75  mS per cm.  This occurred  during  the [x:uk  irrigation  purifxl for cotton, (but see Section 3.4.1).
6LE la: MEAN YIELDS OF SEED COTTON AT MELKA SADI STATE FARM BY
MANAGEMENT UNIT AND SOME RESULTS OF STATISTICAL /XNALYSES
MANAGEMENT UNITS
YEAR                                  1                 2                 3                 4                  5                 6                  7
ANNUAL
MEANS
i '
(QUINTALS
PER HA)
1988
28.1
21.7
20.9
24.6
23.7
22.1
24.0
24.0
1987
31.1
31.9
30.8
33.1
32.2
31.7
35.8
32.3
1986
33.9
31.2
32.1
29.4
32.0
26.4
31.1
31.1
1985
35.5
34.9
31.5
31.8
42.8
30.9
34.8
34.8
1984
37.8
36.8
33.5
35.2
42.7
34.8
38.9
38.9
5 YEAR MEANS
33.3
31.3
29.5
30.8
34.7
29.4
32.9
)ARD DEVIATION
OF SIGNIFICANCE
lRENCES BETWEEN <L MEANS
lRENCES BETWEEN 1 MEANS
6.79 16.1
7.24            8.94            6.50            9.15            6.09            7.71
*
t
♦t
t*
** ♦ t
tT
•ICIENT OF VARIATION 16.1 1.3.2 22.8 13.8 12.35 12.2 14.4
IE: MAAE Office
*     = Significant at 5% level
**     = Significant at 1% level
***      = Significant at 0.1% level
Fields      that      were      abandoned      and      not      productive      were      not      included in these analyses
! lb: MELKA SADI STATE FARM COTION PRODUCTION
AR
- AREA OF COHON
TOTAL PRODUCTION       AVERAGE YTELD OF SEED
r GROWN JN 1IA
OF SEED COTION
TN QUINTALS
COHON IN QUINTALS
PER HA
88
2633
81671
31 .0
87
2633
95108
36. 1
86
2633
103808
39.4
85
2233
91000
40 7
.
7BLE 2: BANANA PRODUCTION AT THE MELKA SADI STATE FARM
CUMULATIVE TOTAL
AVERAGE YIELD
fEAR
AREA OF PRODUCTIVE
BANANAS
HECTARES
AREA OF LAND
ABANDONED
HECTARES
AREA OF LAND
TOTAL annual
ABANDONED
HECTARES
PRODUCTION
IN QUINTALS
OF PRODUCTIVE
LAND
IN QUINTAL PER
HECTARE
1975
720
30
213
171,641
238.4
1976
700
20
233
15,910
271.0
1977
700
0
233
108,050
154.4
1978
733
97
330
106,637
145.5
1979
706
103
433
102,370
145.0
1980
700
190
623
81,353
120.5
1981
700
77
700
80,000
114.3
1982
700
80
780
80,422
114.9
1983   * 280
115
895
37,602
134.3
1984   » 
.. 
400
•
12,433
28.3
1985
440
51,069
116.1
1986
440
260
48,157
109.6
1987
440
23,051
52.4
1988
440
1,155
21,307
48.4
The catastrophic reduction in area between 1982 and 1983 was the result of serious drought in the region and the estate was invaded by local livestock herds for survival. Subsequent re-planting of much of the banana area meant that 1984 production and yield levels were extremely low.
8TABLE 3: MONTHLY RAINFALL DATA FROM STATIONS IN THE ADP STAGE 1 AREA
3a. IAR Agroclimatic Station
1984           1985             1986             1987             1988 MEANS LONG TERM
AVERAGE
January
0.0
18.2
1965-83
0.0
February
0.0
0.0
0.0
0.0
3.G
99.1
8.8
March
37.4
56.6
9.6
38.6
7.0
51.6
April
83.0
167.5
16.5
24.1
58.2
85.5
56.7
May
43.2
95.6
80.5
110.9
39.1
70.3
54.3
91.7
June
43.7
5.7
33.7
44.7
0.0
30.8
July
59.1
0.8
27.G
144.6
21.2
155.6
19.9
107.4
August
36.3
63.3
107.2
123.9
110.2
67.1
106.8
September
136.3
70.0
262.4
130.6
59.5
114.6
135.2
October
34.5
80.1
0.0
71.7
0.0
31.1
November
5.2
7.6
2.0
8.8
0.0
4.3
0.0
19.2
December
0.0
4.1
0.0
2.0
0.4
0.0
12.9
0.0
1.8
1.6
1.3
TOTALS
424.3
490.8
633.0
559.7
685.2
558.4
566.0
3b. PCC Agroclimatic Station
January February March April May
1984 1985 1986 1987 1988 MEANS
0.0               13.2               0.0                0.0              13.8                6.8
0.0
0.0               12.0             45.7            105.2              11.3               34.8 0.0               65.4             39.0              11.2              45.0               32.1 0.0               62.1             11.8              24.0                5.5               20.7
0.0                 3.7             50.6                0.0                6.1              12. 1
127.4             164.5             60.4              26.7              76.9               91.2
42.3               85.4             37.4              G7.2            158.9               78.2
47.9               23.1             61.7              16.7              51.3               40.1 0.0                 1.5               5.3              11.7                9.3                5.6
1.8                 0.0               0.0                0.0                0.0                    3
.0             37.3              19.1              38.0               18.9
June
July August September Oclober November December TOTALS
1.1                 1.6               0.0                0.0                0.0
.7
220.5             432.5           349.2            281.8            416.1              341.5
3c. Pilot Drainage Trial Rain Gauge
January February March Apri 1
May
1984             1985             1986             1987              1988           MEANS 0.0                0.0                O.u               0.0
15.9              10.3               37.3             21.2 32.0              78.7                 9.5             40.1
NO               NO
38.3 16.9
8.9              34.7
June                  RECORD     RECORD              61.7
.0 18.4
.0 14.5
Jul y
.0               12.5             24.7
67.2
62.0
82.6
August
70.6
41.5
69.8
107.4
September
72.9
0.0
18.3
Oc Lober
48.6
22.3
32.6
6.5
0.0
November
13.0
0.0
0.0
0.0
0.0
December
0.0
0.0
0.0
0.0
IQTAI^S
298.1 
297.2 
297.9 
297.7TABLE 4: NUMBER OF PIEZOMETERS BY DEPTH TO WATER TABLE RANGES
ILLUSTRATING FLUCTUATIONS IN GROUNDWATER LEVELS BETWEEN 1984-1989
MONTH
RANGE IN DEPTH TO WATER TABLE IN m.
NUMBER OF         NUMBER CF
OF
OBSERVATION
0.0-1.0
1.01-2.0
2.01-3.0
3.01-4.0
4.01-5.0
5.01-6.0
>6.0
PIEZOMETERS
WITH NO
RECORDS
PIEZOMETERS WITH
RECORDS
June 89
0 (0.0)
14 (22.2)
17 (27.0)
9 (14.3)
5 ( 8.0)
5 ( 8.0)
13 (20.5)
8
63
Jan 89
1 (1.5)
6 ( 9.4)
18 (28.1)
18 (28.1)
4 ( 6.3)
5 ( 7.8)
12 (18.8)
1
64
June 88
0 (0,0)
9 (13.8)
15 (23.1)
10 (15.4)
7 (10.8)
1 ( 6.1)
20 (30.8)
6
65
Jan 88
3 (4.5)
9 (13.5)
10 (15.2)
14 (21.2)
11 (16.7)
2 ( 3.0)
17 (25.9)
5
66
June 87
2 (3.1)
5 ( 7.8)
15 (23.4)
8 (12.5)
12 (18.8)
4 ( 6.3)
18 (28.1)
r.
i
64
Jan 87
1 (1.5)
4 ( 6.0)
17 (25.8),
9 (13.6)
10 (15.2)
8 (12.1)
17 (25.8)
5
66
June 86 Jan 86 June 85 Jan 85 June 84
0 (0.0)          7 (10.9)           15 (23.4)            11 (17.2)
3 ( 4.7)           21 (32.9)                         7                            64
2 (3.1)          5 ( 7.7)
12 (18.5)           12 (18.5)
0 (0.0)          7 (10.9)           15 (23.1)            11 (17.5)
0 (0.0) 6 ( 9.1)
0 (0.0) 8 (14.0)
11 (16.7)           10 (15.2)
9 (15.8)
7 (12.2)
7 (10.9) 7 (10.9) 7 (10.9) 9 (13.5) 4 ( 7.0)
8 (12.5) 3 ( 4.7) 7 (10.6) 6 (10.5)
19 (29.7) • 6 65 21 (32.9) 7 64 23 (34.9) 5 66 23 (10.3) 14 57
( )       % of Total Monthly Observations
10TABLE 5a: ANNUAL SUMMARY OF IRRIGATION WATER SUPPLIED TO THE
THREE USERS IN THE AM I BARA IRRIGATION AREA (CUBIC ME'I’RES X        1000)
YEAR
MAAE
IAR
ASP
*1DTAL
ANNUAL TOTAL
RAINF/ALL IN MM
IAR
PCC
1988
102,910
2,841
5,210
110,961
682.5
416.1
1987
143,092
4,354
12,682
160,128
559.7
282.1
1986
127,531
3,078
12,744
143,353
633.0
349.2
1985
111,874
1,774
7,817
121,465
490.8
432.5
1984
125,094
4,819
12,817
142,730
424.3
220.5
MEAN
122,100
3,373
10,254
135,727
558.4
34 1.5
TABLE 5b: MEAN ANNUAL QUANTITIES OF WATER DELIVERED THROUGH PRIMARY OFFTAKES TO MELKA SADI STATE FARM (CUBIC METRES PER HA)
PRIMARY OFFTAKE NUMBERS
EFFECTIVE RAINFALL
80% TOTAL
YEAR
COTPON AREA
BANANA AREA
APRIL - SEPT
IN CU. M PFT1 ILA
••
1
3
4
2
LAR          t        lVzC
1988
7,940
8,330
10,350
13,438
5,939
3,437
1987
11,145
10,665
7,209
12,234
3,473
1,061
1986
13,181
9,049
9,640
14,751
4,457
2,GO9
1985
13,757
10,421
12,883
22,442
4,706
4,042
1984
11,800
11,596
13,463
15,743
4,090
2,176
MEAN
11,565
10,013
10,709
15,721
4,533
2,745TABLE 6a: RIVER WATER SALINITY AS MONTHLY MEAN
ELECTRICAL CONDUCTIVITY IN mS PER CM
SAMPLING POINT - MAIN CANAL HEADWORKS JUNE 1987 - JULY 1989
1987
1988
1989
MEAN
JANUARY
*
0.40
0.43
0.42
FEBRUARY
t
0.47
0.44
0.46
MARCH
*
0.49
0.42
0.46
APRIL
t
0.52
0.38
0.45
MAY
t
0.67
0.4G
0.57
JUNE
0.35
0.88
0.54
0.59
JULY
0.41
0.55
0.48
0.18
AUGUST
0.36
0.24
-
0.30
SEPTEMBER
0.34
0.24
-
0.30
OCTOBER
0.40
0.31
-
0.36
NOVEMBER
0.36
0.41
-
0.39
DECEMBER
0.34
0.42
-
0.38
MEANS
0.47
0.43
t Measurement began 1st June 1987
12TABLE 6b! ACTHJAL MEASUREMENT'S OF RIVER WATER SALINITY AS Ec IN mS PER CM
1987
1988
Apr 1 0.54
1989
Apr 1 0.37
2
-
2
0.32
3
0.54
3
0.47
4
0.48
4
0.37
MEAN
0.52
MEAN
0.38
May
1
0.52
May 1
0.45
2
0.62
2
0.44
3
0.64
3
0.50
4
0.90
4
0.45
Beginning of
MEAN
0.67
MEAN
0.46
records June 1 
i0.32
June 1        0.89
June 1         0.52
2
0.38
2
0.81
2
0.56
3
0.22
3
0.86
3
0.54
4
0.48
4
0.96
4
0.52
MEAN
0.35
MEAN
0.88
MEAN
0.54
July 1           0.40
July 1        1.04
July 1         0.53
2
0.46
2
-
2
0.52
3
0.38
3
0.34
3
0.43
4
0.40
4
0.28
4
0.44
MEAN
0.41
MEAN
0.55
MEAN
0.48
Aug        1       0.43
Aug        1       0.30
2
0.41
2
0.21
3
0.29
3
0.25
4
0.29
•4
0.20
MEAN
0.36
MEAN
0.24
Sept 1
0.35
Sept 1
0.20
2
-
2
0.24
3
-
3
0.22
4
0.32
4
0.30
MEAN
0.34
MEAN
0.24
Oct 1
0.33
Oct
I
0.28
2
0.34
2
0.32
3
0.34
3
0.33
4
0.58
4
0.32
MEAN
0.40
MEAN
0.31
Nov        1
2        0.38
3        0.38
4         0.32 MEAN      0.3G
Dec 1 0.30
Nov        1       0.38
2       0.40
31      0.46
4l 0.41 MEAN 0.-11
Dec 1  0.43
2
0.30
2       0.43
3
0.38
3       0.40
4
0.38
4       0.42
MEAN1 
0.34
MEAN 0.42
13I
3. 3.1
IDENTIFICATION OF FACTORS LIMITING CROP PRODUCTION LEVELS
Introduction
Production  problems have been  investigated  in  some detail only  on the Melka Sadi State Farm which  can  also  be assumed  to  be representative of that part of the Melka Merer State Farm that is included  in  Stage 1  of the ADP.  The Amibara Settlement Project presents a rather different scenario  since it is not being  farmed intensively  and  production  is very  much  restricted  by  factors which are not limiting  on  the State Farms.  Thus,  because the ASP is a different kind  of operation,  its production  problems have been discussed  in  a separate section.  The Melka Merer Research  Station, the fourth  component of Stage 1  of ADP,  is also  a special case since it is not production  orientated.  A review of some of the results of previous investigations carried  out on  the station  provides some guidelines on some of the factors limiting crop production levels.
3.2 J The Incidence of Soil Salinity
By  far the most serious factor affecting  production  levels of both bananas and  cotton  are the escalating  levels of soil salinity.  In the Melka Sadi banana area land  was being  abandoned  as long  ago  as 1975  and  the loss of land  in  the area allocated  for bananas in  1989 is in  excess of 1200  hectares.  Bused  on  observations in  the field significant areas of these abandoned  lands have bare surfaces where the process of saJinisation  continues.  The actual productive areu of bananas in  1989  has declined  to  little more than  300  hectares whilst the average yield  in  1988  was less than  5  Lons per hectare. Obviously  if nothing  is done to  halt the process of salinisation  of the soil then  banana production  will cease in  the not too  distanL future.
The coLLon  area of Melka Sadi came into  production  in  the early 1980’s.  Since that time a total of 351  hectares had  been  abandoned 
by  1988.  Observations on  the 1989  crop  suggest that this figure is set to  increase since there are a substantial number of fields severely  affected  by  soil salinity.  Average cotton  yields saw a very  marked  drop  in  1988  to  24  quintals per hectare us companx! with nearly  35  quintals in  1987.  However salinity  was only  partially responsible for this dramatic change.
As the f igures for the abandoned  lands in  I lie buianu  area and  the rather pathetic yield  levels of the past few years suggest,  l.hc distribution  of severe soil salinity  is widespread  throughout almost the whole of the buuuia urea.  Thu  distribution  of seven*  soil salinity  in  the cotton  land  of Melka Sadi lias unt il recently  been concentruled  in  ureas adjuccnt to  the buiann  estate on  soils with mixed  textures und  with  relatively  high  waler tables.  However in recent years there has Ix.-en  an  obvious extension  of high  salinity levels to lands far removed from the binaiut. area.
14In  the absence ol' any  routine monitoring  of soil salinity  the pattern  of occurrence is litLJe understood  although  the recent soil survey  has thrown  some light on  the matter.  The most seriously affected  areas appear to  be those where the profile is of mixed tex'tures which  includes the banana area arid  surrounding  cotton lands.  Over considerable areas such  soils are associated  with underlying  volcanic tuff which  comes to  within  less than  2  metres of the surface in  some areas adjacent to  the main  canal.  In  1989  many fields along  the main  canal have large bare patches where high salinity levels have resulted in poor establishment of the cotton.
The soils of mixed  tex'tures tend  to  have relatively  high  water tables of around  2  in  though  not so  high  as to  pose an  obvious threat, in  raising  surface soil salinity  levels.  The lands with  deep 
profiles of uniform heavy  texture seem to  be associated  with  deeper water tables and  are general Ly  not a f Tec ted  by  salinity.  More detailed  observations are required  to  confirm these pre 1  i mi nary judgements.
An  attempt is being  made by  State Farms to  obtain  some aerial photographs of the 1989  crop  and  it is hoped  that these photographs will result in  a more precise definition  of the occurrence of salinity.  Tf saline ureas can  be closely  related  to  soil type and other conditions a more pragmatic approach  to  the drainage problem may  be permitted.  For example soil salinity  may  be related  to  canal seepage,  seepage from the higher ground  above the primary  canal, conditions created  by  localised  variations in  soil textures caused by depositional differences etc.
3.3                   The Causes of Soil Salinity
••
v           Tn       the       first       year       or       two       immediately       following       initial       development of     the     land     on     the     Melka     Sadi     Farm,     yield     levels     of     both     bananas     and cotton        were        satisfactory        and        compared        well        with        other        parts        of Ethiopia         and         the         world.         However         a         relatively        short        Lime        after development        yields        begin        to        decline,        in        many        fields        falling        quickly to      the      point      al      which      land      was      abandoned.      'this      (Incline      was      due      to increasing         salinity         levels.         Tt         is         therefore         important         to         identify the causes of the increase in salinity.
There are two possible causes of the salts which have nccumulal <nl in the top soil to the extent that crop production is reduced or is
not longer possible. Thu salts may derive from the irrigation waler applied or from the groundwater. In both cases l.he salts accumulate in the upper part of the profile lx.cause of evH|X>rnt i on al the soil surface and/or evaporation through the crop, ’these processes remove large quantities of water into the atmosphere lo.iving behind increasing concentrations of salts in the soil water.
15Bused    on    the    monitoring    of    the    Awash    river    water    since    June;    1987, the average electrical conductivity of the irrigation
0.43     mS     per     cm     which     is     equivalent     to     total     dissolved     solids     ol 275       milligrams       per       litre.       The       average       calculated       evapo-Iranspiration of      bunanus      is      approximately      25,000      cubic      metres      per      hectare      pel year,     while     that     for     cotton     is     about     8000     cubic     metres     per     hectare per      crop.      Thus      assuming      no      leaching      takes      place      and      that      only negligible     amounts     are     removed     by     the     crop,     the     annual     accumulation of     salts     in     the     upper     part     of     the     soil     profile     (in     effect     in     the root    zone)    will    be    of    the    order    of    6.8    tons    per    hectare    per    annum for      bananas      and      2.2      tons      for      cotton.      Obviously      leaching      is      a critical      requirement      for      the      removal      of      these      salts      as      the      water table rises into root zone.
The rising  water table has always been  seen  as the major cause of soil salinity  and  this is partly  because a high  water table restricts the amount of leaching  that can  Lake place,  thus preventing  the removal of salts from the root zone that are applied iri the irrigation  water.  At the same time the high  water table loses water in  an  upward  direction  as a result of capillary  suction pressures which  carry  Che water to  the surface.  Invariably  the groundwater has a high  salt content and  its evaporation  at the surface leads to the concentration of salts in the surface soil.
Rising  groundwater levels and  increasing  soil salinity  levels generally  occur becuuse of over-irrigation.  However the accumulation  of salts in  the upper part of the profile can  also occur with  under irrigation  where insufficient water is appLied  for downward  leaching  of the salts.  Following  a brief examination  of water management practices on  the Melka Sadi farm together with  the data on  irrigation  water supplied  it was concluded  that over irrigation  is a primary  cause of soil salinity  through  rising groundwater levels.
The banana area on  the Melka Sadi State Farm is illustrative of the situation  that has developed  as a result of over-irrigation.  This perennial crop  has a very  high  water requirement and  applying  an irrigation  efficiency  of some 70% as is normal for basin  irrigation it is apparent that over the year the quantities of water unused, which  may  be largely  deep  percolution  losses,  are very  considerable.’ Tn  practise irrigation  efficiencies have probably  been  lower than the average for two reasons:
Lhere has Ix.en  little restriction  on  the availability  of water and  no  attempt has been  made to  measure the quantities of water applied, although measuring structures have lx.-cn provided
the        soils generally
i rrigution
of      the      banaria
of
sufficiently 
without obscrvabl
urea of mixed high permeabi J i I e damaging offee
Lux  lures and  ure > I o  |Xii nd  t over Ls in the Ik inarm.Thus in  the absence of drainage the rise in  the water table has been relatively  rapid  and  in  many  areas evaporation  (rom the soil surlace through  capillary  rise from the groundwater has accelerated salinisation  of the soil.  It seems likely  that the cotton  areas adjacent to  the banana lands which  have been  abandoned  have also been  affected  by  high  water tables as a result of over irrigation  ol Che bananas.
The scope for over irrigation  of coLtori is a gocxl deal Less than  Pol bananas particularly  since the cotton  crop  is more obviously susceptible to  waterlogging.  As a result the waler table problem is commonly  not associated  with  the heavy  textured  soils which  have lower infiltration  rates so  that over irrigation  has more noticeably damaging  effects.  Consequently  the incidence of salinity  on  the heavy  textured  soils is less.  The tendency  for over-irrigation  is most acute on  the mixed  textured  und  medium textured  soils where infiltration  rates are much  higher.  Impermeable layers occur in  I he profile of these soils and  perched  water tables arc sometimes found which also contribute to the salinity problem.
An  important cause of over-irrigation  of cotton  and  ultimately  of an increase in  water table height and  salinity,  is the irregular microtopography  on  many  fields.  As a result during  irrigation  the waler becomes ponded  in  the lower spots before it can  pass through the furrow to  the lower purls of the field.  This ponding  commonly leads to  a breakdown  of the ridges with  lateral movement of water arid  a loss of control over application  rales by  individual furrow. Consequently  excessive applications of waler are made which  is unevenly distributed across the field.
Based  on  observations in  the field  it would  appear that a rigid schedule of irrigation  frequency  is being  followed.  It is intended that the timing  of irrigation  should  be determined  by  the soil moisture status as observed  with  the use of auger samples.  In practice this intention is rarely carried out. Irrigation
supervisors have been instructed in the technique of auger sainpl ing for determination of soil moisture status and the subsequent decision on when to irrigate. In practice this technique is nol
used and irrigation applications are made in a routine manner with little or no regard for soil moisture status. It is suggested that
two or three auger samples jx-r field would be the norm in deciding when to irrigate, but irrigators need training in interpret ing (he
’feel’ of the soil. Observations imide in the last three weeks during the soil survey and during other field investigations have
mdicaUd a general tendency for over irrigation at least in (|u younger stages of crop growth und include the following:
soil pits dug in fields of coLton ut 3 - 8 Mwt
.|  stages
demonstrated n profile ut or- near field <:ui»u<:ity to 2 m ni«l these fields have been irrigated n day or- two Inter.
t
have dcpl.l
fields ruinfa I earlier old.
of cotton were being irrigated
1
hail been recorded which hail
by u ruinfull of 31.0 nun. The 17
the day after* 13.0 
beer, preceded two 
col.Lou was 3 - 4
mm ol day j
w< ?ek?fields were being  irrigated  not long  after this rainfall which had  only  small patches of young  cotton  growing.  The remainder of the field  was completely  bare where crop  establishment had failed  due to  already  high  salinity  levels.  The effect of applying  water to  bare land  with  the profile already  at fleld capacity  from previous irrigation  and  rainfall will inevitabl> lead  to  a rise in  the groundwater level and  accelerating salinisation of the soil.
Further evidence of over-irrigation  is provided  from discussions with  irrigators whose main  instruction  is to  ensure Chat Che field is covered.  No  attempt is made to  measure quantity  of water applied and  the natural tendency  will be to  apply  more water than  is necessary.  Night irrigation  which  is an  essential part of the Amibara irrigation  system further complicates control of water application  rates.  Other possibilities for reducing  irrigation  need to  be explored,  particularly  the use of rainfall data for increasing the irrigation interval.
Preliminary  visual observations indicate that there could  be two distinct situations in  which  high  infiltration  rates and  excessive deep percolation losses could occur.
In the first case siltation of minor canals has resulted in reduced 
delivery rates in the canal and reduced flow rates in the furrow, 
which given the long furrow lengths has resulted in increased Lime
to cover the ground and consequently higher infiltration and higher
losses to deep jiercolation.
In  the second  case there are canals where higher bed  and  bank  levels has had  the effect of increasing  siphon  heads with  the result that the flow through  the siphon  has been  increased.  On  unlevel fields water control is adversely affected.
Investigations       are       required       to       conf       inn       whether       these       preliminary observations are a cause of deep percolation losses.
Resource Limitations Water supply
The irrigation water supply for
the Amibara Project Area
the Awash river. In general ample quuitities of water available Tor the irrigation of the total cultivated served by the Amibara canal for the cotton i However an exception to this generalisation beginning of the 1988 cotton season. I
comes     from have
land
river flows occurred through May and June and into July and 
cottori planting programme was much delayed due to the lack of w iter lor the pre-planting programme. As a consequence cotton yield- significantly reduced by the lute planting.
and lianurui i
i occurred
In that year exceptionally Low
I lu* Un-
wereThe 1988  water shortage is considered  to  be an  exceptional event. Normally,  sufficient water is available even  if a substantial amount of water is required for leaching purposes. However the full
development denwind      for there         will quantities.
it is a continual capac ity.
of the Amibara Settlement Project will increase the water. Consequently during periods of water shortage
be more severe problems of distribution of limited Because of serious problems of siltation in minor canals
struggle to maintain the delivery system at
ful 1
Irrigation  water is available from the Awash  river to  support a second  crop  in  the period  December through  to  April over much  of the irrigated  land  areu  served  by  the Amibara canal.  However to  date no advantage has been  taken  of this opportunity  for increasing  crop production.  A major obstacle to  increasing  cropping  intensity  is the heavy  canal maintenance requirement particularly  for de-silting operations which  requires long  periods of cariaL closures.  If this obstacle can  be overcome economically  viable crops and  cropping systems have to  be developed.  At the moment,  the Management arc Cully occupied throughout the year with the single cotton crop.
3.4.2 Machinery and equipment
An  inventory  of the farm machinery  used  on  the Melku  Sadi state farm is presented  in  Table 7.  In  general adequate numbers of tractors and  implements are available to  under take the necessary  operations for the present farming  systems involving  bananas and  cotton. However the target replacement schedule for the farm machinery  is not being  adhered  to  and  a disproportionate number of old  tractors and  implements which  are reaching  the end  of their useful life are included  in  Table 7.  The ageing  tractor fleet and  implements places a heavy  burden  on  the maintenance workshop  and  creates difficulties in achieving timely cultivation operations.
Availability of capital
Funding  is provided  for the Melka Sadi Farm through  the Middle Awash Agricultural Enterprise and  ultimately  through  the Ministry  of’ Slate Farms.  Budgets are prejjared  in  the year previously  un<l in  general re-current expenditure is adequately  funded.  Problems arise with obtaining  funds for capitul expenditure on  major items of*  machinery such  as tractors,  and  for most forms of fixed  investment.  These problems are partly due to the scarcity of foreign exchange.
3.4.4.            Human resources
Based    on    a    very    brief    encounters    with    the it would seem that managerial cuqjacity is
Melka   Sadi   Fa   mi   managers not the major constraint.
19TABLE 7: INVENTORY OF FARM MACHINERY AT MELKA SADI STATE FARM, JULY 1989
MACHINE/IMPLEMENT                   MAKE AND TYPE                  NUMBER HELD
NUMBER
OPERATIONAL
Tractors
130 HP
ti
FIAT TURBO 130-90 DT
7
FIAT TURBO 130-90 Normal
1
100 HP
FIAT 980 DT
3
•I
FIAT 980 Normal
10
IMT 5106 Model
6
75 IIP
II
FIAT 780
24
NTP BE1AURUS
9
TOTAL
53
40
Disc Ploughs
IMCO 3-Furrow AMCO 4-Furrow
ALVAN BLANCH 4-Furrow
BANDAN (BRAZIL) 5-Furrow                       29                      29
Offset Disc Harrows               DUFF
7
MF 3.6 - 4.7 hi
6
Spring Tooth Harrow (not
used)
3
Land Levellers
EVERSMAN 45-12 Model EVERSMAN 329 Model (made in Italy under License)
V
4
Ridgers
RANSOME (UK) 4-5 row
5
RD 23 (Israel)
4
N/tRDT (Italy)
7
16
12
Inter-row
cultivators
Various Models
14
Planters
STANHAY JUMBO Planters
10
7
Trailers
BANGOR (UK)
9
ZIMMAC1I (Yugoslavia)
5
NARDI (Italy)
4
Sub-soi ler
M.F.
3
Mould Board Plough
Make Unknown.. Very large
l-Furrc >w
1
Make Unknown. Very large
2-Furrow
1
Make Unknown. Suuil 1
2-Furrow
2
Rolling cultivator With Lntei-iow
LEHMAN (Li 11 isl.on)
cultivation + Ridger Ditching Mudline/
1
Digger
FOCI AIN
1
20The general appearance of the cotton  crop  in  1989  is impressive and 
behind  the scenes the general standards of record  keeping  appear to 
be satisfactory.  An  important deficiency  of management may  be its
failure to  appreciate the need  for the development of skills at the
Jjel_d_  level.  The basic skills required  Tn"a good  11 au Lui di i veT  ut•
an  efficient irrigator may  not have received  the attention  they 
deserve.  Consequently  unlevel fields and  over-irrigation  have been 
major reasons problem.
for the rapid development of
the soil salinity
The supply  of unskilled  manual labour is a constraint for two specific tasks.  The heavy  work  of de-sililing  tertiary,  quaternary anti field  canals is usually  too  much  for the available labour and the use of mechanical diggers is presently  under consideration.  It has become increasingly  difficult to  attract sufficient manual labour into  the urea to  harvest the cotton  in  a timely  and  efficient manner.  Consequently  the MAAE  have purchased  two  mechanical cotton pickers which will be used for the first time on the 1989 crop.
Husbandry Limitations
1 Bananas
The salinity  problem has become of such  over riding  importance in bananas that other husbandry  practices have made only  marginal differences to  production  levels.  On  the few better fields high levels of fertiliser use have been  applied,  weed  control and  sucker management have been  oT  a high  standard  etc.  However as soi I salintiy  has reduced  crop  vigour,  husbandry  standards have been 
•    allowed    to    slip    and    lack    of    weed    control    and    inadequate    fertiliser use have undoubtedly accelerated the demise of the crop.
• 2        Cotton
Basic cotton  husbandry  practices are good  on  the Melka Sadi fann  as evidenced  by  yield  levels on  some fields which  equate with  those obtained  under trial corudi lions on  the research  fann.  Time of planting  is a major cause of reduced  yields.  Good  weed  control is obtained  by  mechanical method  of control-inter row cultivation  after the pre-plant irrigation  has germinated  weeds is followed  by  further inter-row cultivations at suituble intervals until the colloii height precludes further pisses.  Some manual weeding  may  then  lx*  required to  maintain  control of weed  growth  unl.il the cot.Ion  Conns a Host'd canopy.
Pest     control     is     uohieved     through     aerial     applications     nl     insect.ieidc and is apparently highly effective.
21Fertiliser use is restricted  to  occasional applications ot nitrogen in  the form of urea.  However work  at the blelka Merer I/\R has demonstrated  only  a small responses to  nitrogen  in  a number of trials.  On  the other hand  a long  term fertiliser trial at IAR has demonstrated  that there is no  response to  applications of phosphate and potash even after 13 years of cotton monocropping.
Some improvements could  be achieved  in  cotton  production  by  more careful operation  of ridgers and  planters.  Untidy  ends of the field with  uneven  establishment are common  and  the occasional patch  of unplanted  land  due to  malfunctioning  of the planter could  be avoided.
3.6                The Amibara Settlement Project
This project is being  operated  through  the Relief and  Rehabilitation Commission  which  is a quasi Government Organisation  that does not receive a budget allocation  in  the normal way.  Apparently  it is largely  dependent on  foreign  funding  through  various international aid  agencies.  Thus financial resources available to  the project an* uncertain  and  invariably  inadequate to  enable the project to  make any progress towards self sufficiency.
The project has social objectives which at the moment ap|k*ar to 
over-ride any commercial objectives or indeed any ultempt to fully 
utilise the available resources of land and water. The six: i id 
objectives of the settlement were to accommodate some of the
indigenous Afaa people who lost much of their traditional grazing 
lands with the development of the Amibara project. The participation 
ol the Afaa in the settlement was a Conn of compensation for that
loss and in addition it was an effort to introduce the people to 
arable agriculture, the cash economy and a way of life in which they 
can relate more closely to modern Ethiopian society.
The social objectives of the project have not been  uchieved  largely because of the failure of the project in  productive terms.  This
failure may  be partly  due to  the difficulties in  training  the A fan to  do  manual work  in  the fields and  breaking  the habits of a lifetime and  of the traditional lifestyle in  which  Uh? women undertake all the physicully  onerous tasks.  However the absence of pi open  rewards for the efforts of the Afaa purl icipauls in  the project may  also  have contributed  to  their inudcqiuite p-r forum ice and disinterest.
lhe Ainiliaru Settlement Project lariii is oix*nd.<*d as a collective in which the seLLlers arc required to provide the iiqHits of m.diu.J
a xjur. These inputs are reconled and at lhe end of the season each individual receives a share of Uh? surplus income. However low prtxlucl.i vi l.y levels on the fann has meant that there has Urn little generation of surplus income fur distribution.
22Low productivity  of the projects is due to  a combination  of factors of which  the poor quality  of the manual labour and  semi-skilled workers has already  been  mentioned.  Never Ch  less perhaps the most important limiting  factor has been  the lack  of resources which  has prevented  optimal fanning  practices from being  followed.  The late planting  of cotton  and  maize has been  a regular occurrence due to tractor and  machinery  breakdowns and  the lack  of money  to  ensure the timely  availability  of supplies of seed  and  agrochemicals.  The late planting  drastically  reduces yield  potential and  poor standards of crop  husbandry  subsequently  results is very  low yields. Super imposed  on  this rather dismal picture are the increasing  levels of soil salinity  and  the patchiness in  crop  establishment that is the typical symptom of soil salinity.4. FUTURE AGRICULTURAL DEVELOPMENT REQUIREMENTS
4,1 Objectives and Policies
The Amibara Drainage Project represents a substantial investment by the Ethiopian  Government which  is being  funded  by  a loan  from the African  Development Fund.  This loan  will eventually  have to  be repaid.  Thus it is imperative from the viewpoint of the National Treasury  that maximum benefits are produced  in  the shortest possible time.
The installation  of drainage in  itself will not produce benefits unless the agricultural management is geared  to  take advantage of the improved  conditions resulting  from the drainage.  The identification  of the constraints to  agricultural development in  the situation  following  the installation  of the drainage system and  the preparation  of plans for the removal of these constraints is seen  as a vital component of the Consultants work  prograjiune over the next 18 months.
There     are     well     defined     divisions     of     responsibility     in     the     Stage     1 
area    of    the    ADP.    The    Water    Resources    Development    Authority,    through its    two    subordinate    bodies    the    Amibara    Project    Control    Centre    (PCC) and    the    ADP,    has    responsibility    for    the    operation    and    maintenance    of the    main    and    secondary    canals    and    main    drains,    for    the    design    arid installation       of       the       drainage       system       and       for       various       monitoring activities.       Each       of       the       organisations       involved       in       the       irrigated crop    production    activities    of    the    Stage    1    area    which    include    MAAE, IAR     and     ASP,     has     responsibility     Cor     Lhe     operation     and     maintenance of     distribution     canals     from     tertiary     level     down     to     Che     fields,     tor the      maintenance      of      open      field      drains,      for      lhe      IuikI      preparation operations     and     for     crop     husbandry     from     planting     to     harvest..     Major land     levelling     operations     are     said     to     be     the     responsibility     of     PCC but     this     organisation     has     no     suitable     machinery     tor     undertaking     such work.
The implementation  of the drainage project should  coincide with  a major effort to  improve the agricultural productive pot.enliul and water nuinagement in  the Stage 1  urea.  Tn  other woids a fully coordinated  approach  is needed.  Al the present time there dues not. ap/Xiar to  be any  single mechanism or organisation  to  laciliUit*-*  the coordination  despite the obvious advantages.  Even  al the lcx:al
level Lhe PCC is not
relationship      between      the      slate? a close one.
organlsubtonsThe absence of tangible coordinating  links between  the WRDA and  the Ministry  of State Farms at the top  level and  the slightly  tenuous at the local level creates problems for the Irrigation  Agronomist in fulfilling  his terms of reference most of which  are heavily  involved with  activities at the field  level.  Thus whilst the Irrigation Agronomist is ostensibly  working  for WRDA and  the ADP most of his work  will,  be dependent upon  the cooperation  and  participation  of MAAE,  I AR and  the ASP.  Obviously  his programme of work  should  be developed  together with  these organisations so  that he receives their approval and  support.  In  addition  a clear definition  of the responsibilities for funding  any  field  work  should  be clearly defined.
Positive action  is required  to  establish  a coordinated  approach  in research  and  monitoring  programmes.  For example,  in  research  LAR already  has a programme of research  covering  some aspects of waler management and  including  some leaching  trials and  it has been  agreed that IAR staff will assist in  the operation  of the Pilot Drainage Trial presently  being  run  under the auspices of PCC.  Formal recognition  is required  of the need  to  develop  a joint programme in 1990  of proyierly  formulated  trials that,  will contribute in  a practical sense to  the future improvement of water management on  the state farms.  In  the monitoring  programmes the MAAE already  has an on-going  soil salinity  monitoring  programme being  undertaken  by  an ex-patriate soil scientist in  the Ministry  of State Farms.  This work  must be integrated  with  the programmes that an*  to  be devised for the ADP soil surveyors and  laboratory  technicians presently trained.  A survey  is presently  being  undertaken  by  MAAE  on  the Melka Sadi and  Melka Werer State Farms to  obtuiin  precise estimates of land  areas affected  by  salinity  within  the areas cropped  with cotton.  Such  information  will be extremely  useful in  the planning of priority  areas for drainage and  for preparing  a pirns,  d  programme of development.
Much of the previous paragraphs of this section have Ixien the state farms and IAR because they have muc 1 o <■.on .u 
, loment
early    stages    of    the    ADP.    The    same    cannot    be    said    ot    k=    Setblew    nl Project which is struggling along with minima res > w -
unwilling labour force. The basic requicement wit regard U th< ASP is the need to establish what is to be the policy in tlu future. 
Tn
the
system.
the ]
lands
future change. resources of antici pa Led.
its present state there is no chance of obtaining any return on very considerable investment required to install a drainage
Tf the primary objective of the settlement is to remain ns provision of compensation Lo the Afua piople lor lost, grazing 
» with related social aims then there is little prosptet lor Similarly if the project continues to be sUu-v<d of
equipment and capital then little change can lx?
25A     policy     decision     is     required     on     whether     to     stick     with     the     existing aims     and     objectives     of     the     ASP     in     which     case     the     high     cost     of drainage       could       never       be       justified       economically       or       financially,       or whether     to     invest     the     capital     and     provide     the     management     which     could transform       the       project       into       an       economically       viable       enterprise       tliat could      justify      the      investment      in      drainage.      It      should      be      pointed      out that     the     ASP     represents     valuable     resources     of     land     and     water     which the country can ill-afford to remain largely unutilised.
The re-development of that part of the ASP which  has been  allowed over the last 5  years to  revert to  heavy  Acacia type bush  has been discussed  and  it lias been  included  in  the Stage 1  area of ADP. This area is presently  used  by  the indigenous Afaa as a grazing  area and  its re-development will be a loss to  those people.  If the chosen  policy  option  is the full development of intensive irrigated agriculture on  the ASP then  clearly  some irrigated  pasture and/or forage crops should  be included  in  the rotation  to  be made available to  the Afaa livestock  as a form of compensation.  Indeed  the supply of livestock  feedstuffs to  the Afaa may  be a better*  way  of encouraging  them to  undertake manual work  on  the arable crops Ilian the present profit sharing system.
Land Levelling
Background
No  major corrective levelling  has been  undertaken  since the original land  development work  was done on  the project prior to  planting  the inaugural crop.  Poor*  land  levels are inevitable however*,  due to some consolidation  through  irrigation  making  secondary  levelling necessary.
Since    development    six    or    seven    cotton    ciop»    *    there    has    been    iunple
that      even      if      the      levelling      work      was      well      uo
sui( e
.       that      initial
opportunity      for      the      disturbance      of      land      eve      .      (llkcr()Lol
x)graphic      and 
operation. An important factor* creu 
sometimes macrotopographic variation has -• operation, particularly with disc plouguns-
Some tractors dr ivers were properly 
irrigation project but th a cocnplete h
quickly lost any pride they had in a
mounds und untidy f inishing 
not been improved in recent y<:n«» 1
untrained and unskilled tractor <>Vo correct the •• ' i»uv<* deteriorated despite the ‘-1 
improper machinery
.
mrcessnry * 
the early days of the of incentives many 
done uncl furrows and The si Lua( ion has
(Miiploymoril of
Consequent!v land levels
.... . 
llin d evolleis. 
o|*
shoddy pl oughuiunsh i p with use o hveisnui lllUg or smo<>
|Ull |
,y
this uLhine is designed 
requires a careful experienced iqx.ru resulLs.
J? achieve sat i st a to..Unlevel fields have many  adverse and  costly  effects,  particularly  in fields where the length  of run  of the furrows commonly  exceeds 200 in. Some of these effects are briefly outlined below:
i.       Depressions    in    the    field    cause    a    ponding    of    water    which    can, and    often    does,    lead    to    a    breakdown    of    furrows,    the    lateral movement    of    water    across    furrows    and    a    waterlogged    patch    of crop.      Growth      of      the      crop      is      severely      retarded      from      the seedling stage onwards resulting in lowered yield levels.
ii.       The ponding                of water in depressions eliminates
any
possibility      for      control      over      the      quantities      of      water      applied either to individual furrows or to the field as a whole.
If the field  is to  be covered  with  water the application  nite is invariably  excessive where the land  is not level . Irrigators are instructed  to  ensure that water reaches all parts of the field  and  the ponding  of water that is inevitable iri carrying  out this instruction  results in  very 
low field irrigation efficiencies.
iii. 
The low field application efficiencies on unlevel fields are
caused by excessive losses of water to deep purcolution and to direct evaporation from the ponded areas. The deep percolation losses contribute to increasing soil salinity levels both at the surface and down the profile.
iv. 
The    unlevel    fields    have    been    observed    to    increase
the
labour
requirement.    Two    or    three    men    may    be    required    to 
unsure
that
complete cover of the field with irrigation waler
is
achieved.      The      additional      labour      is      necessary      to 
lonn 
in
field    bunds    to    prevent    lateral    movement    of    water
arid    to
re
form and maintain the ridges.
v.     Over Lime the low field application efficiencies inevitably lead to u rise in the groundwater level which il unc iecu. . Y any form of drainage, will eventually reach a i.vi. a. w i a water will rise to the surface by capillary suction. >: evaporation of this water at the soil surface rusu sin 
ever increasing concentration of salts in the up[x r P* . 
the profile with increasingly observable adverse el leets crop perfonaance, to the point at which it U.-cdues no lony.r viable to continue cropping the land.
vi .      The initial signs of salinity as
to establish the coLton crop urn -
demonstrated by Lire failure * c<MiMiiunly related l.o areas of
-
small        depressions        in        the        land        surlace.        *sc        1
observation it may be concluded that lb’-
°
•* levelling has contributed to more rapid sa
soi 1 than would have occurred on                                                  1 conditions     in     which     wuter*     application     ra.es     c> careful ly control led.
j
Illis land 
’   l.h<- under more
214.2.2            Extent of land levelling requirements
It is apparent that many  fields in  the Melka Sadi cotton  lands and elsewhere in  the cotton  lands of the Stage 1  development urea are unlevel and  that irrigation  efficiencies an? drastically  reduced  us a result.  Arandom study  of this was carried  out during  the Master Drainage Plan  Study  \1.  Nevertheless no  detailed  studies have so far been  undertaken  to  establish  the requirements for major corrective land  levelling.  Without such  studies one- can  only estimate the size and  scope of future land  levelling  requirements in the Stage 1  area of ADP or elsewhere on  the MAAE  farms. Consequently  priority  should  be given  to  the studies required  so that once the size of the problems luis been  determined  an  action plan can be prepared on how to tackle the problem.
4.2.3 Option for land levelling
The option for undertaking the land levelling work Kus now decided. The equipment will be procured through ADP I ;uU Kindid over to PCC who will carry out the land levelling on b-kill ot the benefitting entities. A charge to cover costs will U* madt*»
4.3               Land Reclamation
The main requirement for the reclamation of aKuuh a Stage 1 area of ADP is for the rental of
the profile. 
The installation of an el fcx.ti
•
required    before    the    leaching    process    for    the    ivu>o    a cun Luke place.
Leuching and the removal of salts in the effective action that cun lead to the full ~ abandoned areas. However every efloi
the continuing increase in su Unity whi< > already abandoned and in lands which will l"T year or two.
‘^wl uf pn^ntl.v to  ‘
1  kind uliu e in lands , in the
The on-going accumulation
in ureas made devoid of — 
of    suits    al    the    soil    surface    arises    tan    rd)
vegetal .ion by cultivation ;u*l when- sal I
. cSlabl ish'* ^
1
tolerant vegetation has not yet " *  . ^d in<
i
In these an a* 
water tables are at a level fro™ whuh suU
the surface by capillary action him *
leads to increasing concentrations o concentration in the upp‘>* l*  rules niv much
__ .* gixxuiduutrr irnivcs to 
•viqiorul io<i from the haw sails
... 
» Uu.
u
. 'Ilk- li ighrr the sal I prof i lr the twin* s*»i I
structure breuks down, infills ?. \ • 
1‘Mudiing requirements in tonus o
l    l       iunl quantities
28Tn  order to  halt this process of salt accumulation  at the bare soil surface every  effort needs to  be made to  establish  salt tolerant vegetation  both  on  fields already  abandoned  and  on  fields that are newly  abandoned.  The local star grass,  which  appears to  have good tolerance to  saline conditions,  is easily  propagated  vegetatively and  has good  nutritive value for livestock,  seems worth investigation  for this purpose.  There will be no  problems with  the removal of star grass in  conditions in  which  the soil moisture condition  is controllable for much  of the year.  The vegetation  will assist the reclamation process in the following ways:
it will use some water from the groundwater table and  most of the rainfall and  will therefore halt the rise,  and  may  even resul t in  some reduction,  of the level of the groundwater table.
it       will       maintain       the       physical       condition  of the soi1 greatly       improve       infiltration       rates       and       ultimately leaching.
ttheherree i iss ge  genenerraalllly s y somomee upt    uptaakeke of  of s saallttss f frr omom t thehe s soioil
tolerant     plants     which     in     itself     may     assist     in     the soil salinity levels.
the
and so 
rale of
f
reduct
ion 
As the method < need for Obviously
------  ----------
might fulfill this need, but no worc and requirements until the drainage - of salt tolerant crops should be coordination meeting in February
installation of the drainage system proceeds a practical ° eaching will have to be developed which may include a
he planting of salt tolerant crops in the early years, trial work should be instigated to identify crops that
! done on leaching methods is installed. Planting 
during the* first major
4.4
Improvement of Water Management Prac The deficiencies and limitations . ry|  prcpim »°
|e
l
p oresent wd’C . 
practices have been discuss cal
fields will greatly enhance the ^7control over q—tU Hianagement. With cotton the
water applied to the fields wi K-
for 1 lUli ii.ies
. improved,
UU inag^»ucuL level
improved water
.      ‘ ‘       ; of
I ri order to lake
advantage ~
Irrigation scheduli~------- ing is required. At the present time the
to be firmly f] <d. hittie account u
°f this more precise control a more
. lS tid<en
very heavy falls occur. 
f; rop ure not reflected in lower scheduling should be determiiu-d oy obser-vution which might include.
a.
x
l°w<J
the theoretical calculation 
•standard assumptions on soil moi. depths, effective rainfall etc.
29
pragmatic npprmu.'h l.o
being followed appear
-     — of the ruinfall unless
water requirements of the young<?r application rates. Irrigation 
a combination of cnlculul ion /uni
of (jorisumj)IJon use appli<.*<l L<i ‘ characteristic's, rooting
• r -■b.  the observed  soil moisture status of samples obtained  with  the use of an  auger,  or by  the use of gypsum blocks on  selected areas.
c. 
the      observed      state      of      the      crop      during      period      of      maximum temperatures.
AL present the use of gypsuin  blocks for determining  soil
moisture      status      is      being      evaluated 
encouraging. 
Laboratory 
analysi
s
by 
is
TAR
also 
and      the possible
results
for
are these
tests at IAR.
d. 
factors
limiting       pro
duction 
such 
as
weed 
control
should 
bt
considered. (S
ee Appendix 1).
Quantities of water applied can be accurately assessed from a
calculation involving the number and sizes of syphons used, 
approximate head and the time which they allowed to flow. However, 
a system of measurement of water application rates cannot be
introduced       in       the       present       circumstances       for       two       important       reasons: the unlevel fields which have already been referred to aid secocdly
the general luck  of expertise amongst management,  the Held 
supervisors and t.ho 
irrirfufArc Time ImvifW l^v*dled the
4.5 4.5.1
vupauie field irrigators the work shouLd be
semi-skilled and should be rewarded according •
-St ions thut must be faced with Lhe banana crop are:
301.5.2           Cotton
The       problem       with       land       preparation       practices       in       cotton       does       not       lie in      the      techniques      employed      but      rather      in      the      way      in      which      they      are carried        out.        Correct        operation        of        tractors        and        implements        is        vital in       order       not       to       cause       uneven       land       surfaces       and       subsequent       adverse effects         on         irrigation         efficiencies.         The         proper         training         of         tractor drivers         is         an         essential         first         step         towards         correcting         unconventional practices                but in addition proper rewards must be given for
recognised skills and for a job well done whilst at the same time
introducing disciplinary procedures to pull those drivers into line
who do not perform satisfactorily.
Alternative systems of land  cultivation  may  also  be worth investigation.  For example the use of chisel ploughs would  provide a depth  of tilth  similar to  that of the disc plough  for easy breakdown  by  disc harrowing  whilst at the same time creating  minimal disturbance of land  surface levels.  The chisel plough  may  also  be appropriate in  so  far as the actual speed  of the soil cultivation  is not an  essential part of weed  control and  it will give a somewhat higher work rate than the disc plough for the same tractor power.
Time of planting  is a critical element which  determines the potential yield  of cotton  in  the Middle Awash  and  indeed  elsewhere
in  Africa.  The optimum period  in  Ainibara,  as demonstrated  by  trials undertaken  by  IAR,  is from early  May  through  to  mid-June.  Yields decline rapidly  with  later planting,  the rate of decline being proportional to  the number of days outside the optimum period. Although  the importance of time of planting  is fully  appreciated  by management circumstances usually  prevail which  result In  late planting of at least a part of the crop.
Tn 1988 responsible corrected, significant part after mid-June, difficulties which the farm machinery
lack of water for the critical pre-plant irrigation was for delayed planting but such a deficiency cannot bo 
1989 delayed planting was lessIn 1989 delayed planting was less serious but a
of the crop on the two State Farms
In c/in and
this year delays were due
be corrected by improving the management efficiency.
Ferti1iser restricted of* around
use on       the State Farms at Melku Sadi und Melku Werer
was planted to operational
re1i abi1i ty o f is
to     the     appl     ication     of     urea     us     u     top     dressing     ut     the     rate 100 kg per hectare (45 kg N) to cotton crops which are
yellowing,  slow growing  and  lucking  in  vigour.  Tn  some fields where the cotton  growth  is not vigorous it is difficult to  distinguish between  the effects of soil salinity  und  nutrient deficiencies.  It may  therefore be useful to  investigate cotton  yield  response,  to applications of nitrogen,  phosphate und  potash  on  these aretus whore the cotton  is less vigorous which  is usual ly  associated  with  the lighter textured soils. Previous fertiliser work uL I AR has been
carried out on vertisol soils.
31Pest        control           operations are largely restricted to aerial
applications     of     insecticide     based     on     a     scouting     system     with     some spot       spraying being carried out using knapsack sprays where
appropriate.  Insufficient observations and  investigations have been undertaken  to  express an  opinion  on  the effectiveness of the pest control programme.  However one observation  is worth  mentioning.  In mid-July  large numbers of dead  fish  were observed  in  the main  drain which  it later transpired  was due to  the careless use of insecticide either from the aerial application  or through  washing  manual spraying  equipment in  the drain.  Pesticide use should  be kept under continual review and  every  effort should  be made to  prevent damaging environmental pollution outside the cotton fields.
Agronomic Trial Programme
Much closer links should be forged between the State Farms and the ° <a Werer IAR. Meetings were held in May 1989 when an outline
wor t plan (proposed by TAR) for trials to evaluated surface irrigation, pre-irrigation of cotton and wheat, :unl a study of crop water requirements. Representatives of State Farms were present at
re meeting and it is important that the latter should have some say in determining the IAR programme of work and off-station trials.
years and
Banana production has been in steady decline at Melka Sadi for many yet little research effort has been made to investigate
possible approaches to halting this dec liie.
crop  is almost al the point of cxtmc ion 
As  a  result  the  Umana
and there
indication i without drainage.
as to whether sustainable production is poss
‘ is liLtle ssi ble with or
With a production
cycle of a• t. leas• t 5 
%           1 
•*•5 or 6 years
it takes years to obtain results from any banana
Hndep theto dever°ni,nen^ ^ conditions of the area it is likely that the
field experiments.
-- -
. I < _•            a
a
^^ elop
v
ll(in^
sus tainable banuna production lies in the
^^^icient dr-'  **  SUltable rotation,  assuming  of course that an 
r ‘fe£d  |«o linage system is installed.  Other problems that will
\ ^ation t
rr
I ’^querioy 
r
after drainage include method of surface use ^> irrigation water application rates and
• fertiliser use, etc.
^fe^ching t ei   riaLs will need to lx? conducted following drainagf
r<?
duevd by ?
0    L°
the rules at which salinity levels can
f<Jr vnri Un°Wn Huiuitities of water, on Lite various soil types mid c ‘ulhnn Hr. editions of salinity covering both Hl- Uumm and 
tin. 
this resjx-d reference should also U- s oi| Uru] Groi.mdwal.<?r Monitoring Pix>gmii«*‘ f(,r
to
A.The results of these trials will be used as the basi^ fnr „i ■
leaching programme for the Stage 1 area and for restoring fhTSi
to a condition m which the traditional crops of bananas and cottu 
can be planted.. Alternative salt tolerant crops need to be trie
J
operational.  Preparations for the leaching  trials can  begin  with the identification  of suitable sites representative of the range of soil conditions within  the Stage 1  drainage area.  Infiltration tests might usefully be carried out at these sites.
Agronomic trial programmes for cotton  should  encompass cultivation trials,  fertiliser trials,  and  water management trials.  Subsoiling for cotton  will only  be important where there is a plough  pan.  The cultivation  trials,  which  are seen  as low priority  because of the rarity  of soils with  a plough  pan,  are required  to  determine the effects of deep  cultivation  and  sub-soiling  on  cotton  yields. Fertiliser trials are also  of low priority  but could  provide useful information  if carried  out on  fields of known  low productivity. Deep  cultivation  may  be worth  investigation  to  see what advantages are to  be gained  from reducing  the salt concentration  in  the top soil.
Water management trials and investigations are seen as luiving the highest priority and the requirements are discussed in more elai
in Section 5.
along the sides.
Training
this report including the water management drivers. Practical instruction is require
organised and completed before the completion* ^ork which nrotvd.lv best follows the mstalla
||rseS should U'
k vc
.
lling
d|ll iiu<gr.
and PCC.
33Amibara Settlement Project Planning
Section 3.6 of this report highlighted the problems associated with 
the Amibara Settlement Project at present and the failure in real
terms of both the social and the commercial objectives. In 
particular the view is given that the social objectives have not
been achieved because of the failure in productive terms.
During    the    course    of    the    preparation    of    the    Master    Drainage    Plan    for Melka      Sadi      and      Amibara      Areas      \1,      the      Relief      and      Rehabilitation Coiiunission     proposed     changes     to     land     use     and     settler     land     allocation within     the     Amibara     Settlement     Project     although     at     the     time     this     did not       represent       Government       Policy.       The       infrastructure       and       cropping pattern     was     to     be     re-organised     to     accommodate     extra     families     caused     by the      proposed      acceleration      of      family      settlement      and      partly      to      make the     area     self     sufficient     in     food     and     fuel.     The     proposed     land     use ratios were as follows:
% of Area
Cotton 
Maize..............................................................  • Vegetables
14
14
$5
Pasture (for the Dairy Enterprise)
5353
Forestry
During the report review varying land use plans were also evaluated. These were:
A reduction in the area of the dairy enterprise
An alternative cropping pattern with 100% cotton ) An alternative cropping pattern of 50% cotton, 25% maize
25% pasture.
14
Hul
crow          Pur triers, July, 1985Plans could  then  be prepared  for the developnent of the physical infrastructure covering  the irrigation  and  drainage systems,  the field  lay-outs and  other on-farm infrastructure.  Given  the selected  crops and  the chosen  farming  system,  the requirements for on-farm equipment,  buildings and  farm machinery  would  be determined and  cost estimates prepared.  Crop  budgets would  be prepared.  Human resources required  would  be estimated  including  managerial expertise and  skilled  and  manual labour needs.  Full account must be taken of existing resources on the project.
A    detailed    phased    developnent    programme    should    be    prepared    with    a corresponding      schedule      of      capital      investment      requirements.      A      cash flow    analysis    could    then    be    carried    out    to    provide    the    controlling agency    with    sound    estimates    of    the    financial    resources    that    have    to be made available to the project.
Economic and  financial analyses of the proposed  project development should  be carried  out since they  will be of particular importance if funding  is to  be sought from outside sources i.e the international banks or aid agencies.5.                  PROPOSALS AND RECOMMENDATIONS - ACTION PLAN FOR 1990
5.1 5.1.1
Project Monitoring Improvements and Innovations
Soil salinity monitoring
This aspect of monitoring  although  of critical importance in  Lhe Ami barn  Project Area has been  almost completely  neglected  in  the five years prior to  1989  and  the only  significant data available from this period  is from the pilot drainage trial.  The latter is of
only     limited     value     in     understanding     the     broader     scene.     However     1989 
has seen  a flurry  of activity  in  soil investigations.  The ADP soil survey  has provided  some? bencluiuirk  data over some 1100  hectares and a soil Scientist from Lhe Ministry  of State Farms is said  to  have taken  more than  300  soil,  samples on  the Melka Sadi State Farm. Obviously  the data gathered  from these two  surveys should  l)e pooled for mutual benefit.  Similarly  the data held  by  the TAR should  be freely exchanged with the project.
Although  the results of past surveys provide good  information  and are useful baseline data,  future soil suliniLy  monitoring  must have a more specific focus and  be related  to  precisely  defined  conditions if it is to  lead  to  a better understanding  of the build-up  in  soil salinity  levels and  its effect on  Lhe growlh  of crops and  Hit? natural vegetation.  The monitoring  should  consist of measurements of the electrical conductivity  of Lhe soil extract solution,  1:2  by volume,  and  if present measurements demonstrate that sodium is a problem in some areas Chen sodium levels should also be monitored.
Some     proposals     for     Lhe     priority     areas     in     which     Lite     monitoring     of soil salinity should be focussed are broadly outlined below:
comparisons      of      salinity      levels      of      soils      within      fields      from saline      patches      where      cotton      has      not      germinated      and      from adjacent     healthy     cotton.     This     situation     occurs     on     lands     within the    Stage    1    drainage    area    on    both    Lhe    Melka    Werer    and    Melka Sadi      State      Farms.      Monitoring      of      selected      fields      should      be designed     to     cover     both     the     pre     and     post     drainage     si     tuitions acting     as     a     measure     of     the     effectiveness     of     leaching     following the installation of drainage in reducing soil salinity levels.
comparisons of salinity levels in the soils ol the abandoned banana lands including fields completely devoid of vegctaLion, fields with a natural grass cover, und fields with natural bus) regeneration and tall, grass cover. This soil monil.oHng won » also take into accounL Lhe depth to groundwater. lhe major objectives of this monitoring would be determine the i.d.<
(>
regeneration     in     Lhe     three     sets     oi     conditions     and     to     piowo buscline data representing the pro-drainage situation.
correlation of soil salinity levels und cotton
yields. Ik used 
<'i* Lhe selected cover a
yield daUi for the past 5 years, lbdds wj U£Hiid for soil salinity monitoring und these I odds range of production levels - high, m».t iuiii ai
36These selected  fields would  also  be examined  to  detemine water- table depths and  salinity  levels,  soil textures in  the profile to maximum rooting depth, and water application efficiencies.
The monitoring  of soil salinity  levels should  be concentrated  on  the upper part of the profile.  Sampling  below the rooL  zone of the cotton  should  only  be carried  out if it is clearly  of importance -
for    example    in    the    case    of    high    water    tables    where    the    groundwater is saline.
5.1.2 Aerial photography
The simplest means of monitoring  the cotton  areas affected  by  soil salinity  would  be with  the use of aerial photography.  Thu  month  of August or early  September is probably  the most appropriate Lime.  A serious effort should  be made to  investigate the possibliti.es for talking  photographs on  a regular basis,  probably  annually,  as measurable changes appear to  be talcing  place fi-om yeur to  year. Perliaps this could  be arranged  using  the planes used  for pesticide applications or possibly  through  the military.  Tl is understand that MSFD now intend  to  organise the talcing  of regular aerial photographs.
1.3 Monitoring water quality
The present monitoring  of water quality  is restricted  to  weekly samples taken  at the headworks,  the main  drain  and  al the drain out.let into  the river.  Future sampling  should  be expanded  and increased  to  cover the drain  water quality  across the Stage 1  urea. Groundwater has been  shown  to  be of highly  variable quality  across the Stage 1  area and  it would  be of interest to  know whether these difference were reflected in the drainage water.
Siltation  in  the canals of the project is a serious problem.  The information  available on  silt loads is very  limited  but some
•noniLoring   had   commenced   in   July   1989   and   should   Im.*   continued.   Hie objectives    of    such    monitoring    would    be    to    identify    the    need    for    more careful     water     management     to     reduce     the     rule     of     siltation     and/or     the need for additional silt traps.
5.1.4
Monitoring crop production levels
A more systematic approach is vequl1 production         data         particularly         for including date of planting und trumb ‘
should also be included. This
•   -   <   in  other relevant tan
the   recording   of   Hyhl
(O U
;
\ djUvS             of         irrigations, 
ful b lussvS
s.x ux
used us a management tool. !«' J-’8
were planted to cotton which with a p ' 
y ield data would have 1-een ulxu.doued. been cultivated and the patchy eslal.l .s I ... n . that large ureas of Uu-e soil have
to raise groundwater levels und lesu in salinity of the soil surlnru.
37
l?^bh- nunb- of I - ds “ ’ . investigation ol P'1 v 11
• • ‘
-a()
 these I ieh s •
col Ion has n. •        • d
lllg  .a.,.—Monitoring  of crop  production  levels will also  provide baseline data for the measurements of benefits arising  from the installation  of Lite drainage system.
Definition of Size and Scope of Land Levelling Requirements
During  the period  following  the removal of Che cotton  stalks of the 1989  cott.on  crop  when  the first lund  preparation  operations are taking  place,  detailed  topographic surveys should  be carried  out to provide a more accurate assessment of the quantities of soil that have to  lie moved  in  the land  levelling  process.  These surveys should  cover a range of conditions and  should  include both  cotton and banana areas.
The latter are of lesser importance because furrow irrigation is
used only in the early stages of banana irrigation lie fore changing 
to the ljusin method.
5.3               Water Management Observations and Measurements
There are some obvious indications that waler use efficiency  in  the field  is at low level and  that percolation  losses below the root zone oT  the crops are a major cause of Lho  rising  groundwater level. The most important factor responsible fur low field  application efficiencies is the variations in  micro  topography  but there are a nuiiitier of limitations due Lo  poor water management practices which are so far ill-defined.
Tn  order to  define the water management limitations on  Held application  efficiencies,  investigations are required  into  the practical operation  of the systems of water allocation  and irrigation  scheduling.  Observations need  lo  be carried  out into present field  irrigation  practices,  although  preliminary  work  has already  been  done on  this before.\l Measurements ol actual water use should  be made so  that accurate estimates of existing  water use efficiencies can  be provided.  This will be reviewed  during  tin. visiL of the Consultants Water Management Engineer early in 1990.
Tin? furrow length of run in most fields is al least 200 m and in a fields may extent up to 300 m. Slopes are very flat there is a
need to investigate field application efficiencies under various s °il conditions and slopes given these long furrow lengths.
IAR. 1989
38
-5.4
Observation on Groundwater Fluctuation
The observation  well data which  has been  gathered  during  the last five or six  years do  not reflect in  many  cases the expected  or predicted  rise in  the water table.  Tt would  be useful to  determine whether the observation  well leadings taken  within  the Stage I drainage area are a true reflection  of the actual depths to  water table in  surrounding  areas.  For this purpose the installation  of a number of piezometer or auger holes around  each  selected  observation well is proposed  and  this should  be carried  out at the earliest opportunity  provided  that a budget allocation  can  be made to  do this.  These piezometers will not only  serve as a check  on  the observation  wells but will also  given  a much  better understanding  of the fluctuations in  water table levels which  according  to  some observation wells may reach us much as 2-3 m.
5.5               Applied Research
5.5.1             Leaching and land reclamation
In  order to  undertake leaching  and  land  reclamation  trials it is essential to  have a drainage system that will remove the saline drainage water.  There are no  suitable new sites fur such  trials until the drainage system has been  installed.  However in  the meantime the maximum amount,  of information  should  be extracted  from the on-going Pilot Drainage Trial on the Melka Sadi State Fann.
The initial step  on  the Pilot Drainage Trial should  be to  undertake a detailed  analysis of all existing  data that luis been  gathered  over the three year life of the trial.  Based  on  u  cui'sory  examination of some of the data it is apparent,  that there are many  deticiencies und  some records are of doubtful value.  In  addition  there tire iwiny Weaknesses in  the experimental techniques employed.  Nevertheless, it may  be possible to  salvage something  from the available data even if only broad trends.
395.5.2
Correct      experimental      procedures      in      the      Pilot      Drainage      trial      which could      be      followed      for      the      1990      cotton      crop      should      be      prepared. These procedures will include:                                      the method to lx? used for calculating        application        rates        (including        the        leaching        requirement) and       irrigation       frequencies;       the       measurement       of       the       quantities       of water     applied     to     each     plot     and     the     quantities     of     water     drained;     the clear     definition     of     plots     (net     and     gross     areas);     the     crop     husbandry practices     to     be     followed;     the     data     to     be     recorded.     To     date     these ar<i not results which cun be upplied to the main crop.
Similarly  for the banana area of the Pilot Drainage Trial experimental procedures should  be introduced  which  will enable meaningful results to be obtained.
Bananas
From the mid-1970’s there has been  a steady  decline in  banana yields and  there has been  a regular abandonment of land  previously  cropped with  bananas.  The point has almost been  reached  at which  Che decision has to be made to close down the banana operation.
Unfortunately  banana research  has been  much  neglected  in  the past and  little attempt has been  made to  investigate the causes of declining  production  levels.  The assumption  seems to  have been  that soil salinity  was wholly  responsible for declining  yields and  that the effects were irreversible and  uncontrollable.  This assumption should  be tested  by  planting  a series of observation  plots on  lands that have been under a bush fallow for some years.
Bananas are known  to  be susceptible to  a range of soil borne debilitating  diseases including  various wilts,  nematodes etc and have a high  nutrient requirement.  Thus the advantages of rotational cropping  are worth  investigation  even  in  the most favourable environments.  The Melka Sadi State Farm provides conditions which are far from ideal for the banana crop  and  to  this must lx? added  the salinity  problem.  The combination  of an  unfavourable climate with saline soils suggests a need  for rotational cropping  if favourable conditions are to be maintained.
Pollowing the installation of an effective drainage system a long banana rotational trial should be established to examine Che
y <-cts of various cropping sequences on the sustainability ol Anana production. Rotational break crops would include pasture, 
n manure crops, and several annual crops, in both suiumet an winter seasons.
405.5.3           Cotton
The baa io land preparation operations presently undertaken fo 
c*****        ot ton crop i~—- s1 eem to be perfectly adequate providing good 
n ci iC P4           *• J—*•*- - ** *  v------- JHUV1U
conLrol and a sufficient depth of tilth on which 
ken  for the
___ I weed
satisfactory size can be prepared. Z_ 
-       -w.1       ridges       of
a
can    be    prepared.    As    far    as    can    be    ascertained 
there are few areas where pan formation has reached the degree
compaction at which rooL development and water percolation iuijxxled. However a trial to investigate/ the effects of sub-owllm
of
are
in a field in which there is a definite plough pin {or impermeabl
layer) is considered to be worthwhile. e a Ll_. become more extensive as the years pass and the free
•-soiling 
le
The pun problem is likely to 2          ‘ movement of
even 
greater
depth 
than 
the
disc
plough 
and 
at
relatively 
greater
speed. 
Tt
is
recom
mended
that
a
suitable
type
of
chisel
plough 
5.5.4
water     through     the     profile     to     facilitate     systematic     leaching     of     salts is most important.
Tlie min  problem with  land  preparation  at the present time is thuL land  surfaces are frequently  disturbed  by  the disc ploughs as presently  operated.  Surface disturbance is caused  by  the furrows of disc ploughs und  poor finishing  resulting  in  unlevel fields and water distribution  difficulties.  It is suggested  that a chisel
plough  would  disturb  the land  surface much  less than  the disc plough leaving  no  furrow.  A suitable chisel plough  can  cultivate to  an 
should      be      introduced      and      tested.      Investigations      are      proceeding      to identify suitable models and to obtain prices.
The field observations made during July 1989 suggest that the frequency of irrigation is generally too high (i.e. the interval is
Loo short) and that little account is taken of rainfall. However, over i rrigation during low demand periods (April and May) <ind under irrigation during the rest of the growing season has also been 
reported.\K In order to develop a method of rule of thumb assessment of irrigation need, trials should be undertaken to investigate the effects of irrigation frequency on yield levels to supplement those already carried out at IAR.
Miscellaneous trials
The terms of reference refer to the
of the introduction ol a second crop { • • 
chopping) into the limitations ol
nianagement maintenance and
farming      system      of      the      State      Fumis.      OWlV^      rfeauired      for      canal
and machinery, the long 
the lack of any experience
suggest LhuL in the short te—   rm          -
all suggest Lik*^ *»• 
realistic possibi I ity. Nevertheless a trial y >
in which the range of potential c-rops a
the long teem future. Thus* al ter na a <. 
«. possible break crop. i»
be (frown us part of the rechuimLiun pr
G -HaiJer, Melluimu Ri-gca, ETII/82/014 Field fkx;uu<:nL No.
■11
emps
double vropp^J*^.;;
Fri-il programme
_____xiuniiicd  with  a view to '‘crops may  also  have a rob 
•
and/or crops
I
J“'‘
\1Some applied  research  should  also  be undertaken  to  investigate ways of Improving  die income levels generated  on  the Amibaru  Settlement Project.  This would  involve trials to  test the adaptation  of potential crops to  /Unibara conditions and  assess their productive potentials.  Some crops already  grown  in  /Vnibara village have potential and  these include sugar cane,  papaya and  citrus.  The development of sugar cane for the production  of jaggery  as a
•cottage industry’ has possibilities.
.5.5 Implementation of a crop trial programme
There is an  important need  when  formulating  proposals for an  applied research  programme for a co-ordinated  approach  in  which  all interested  parties are involved.  The Melka Werer IAR has obviously the key  role in  the implementation  of an  applied  research  programme but the MAAE  and  the State Fann  managers and  the Amihara Settlement Project managers should be involved in the definition of priorities.
Before any  proposals are formulated  a review should  be undertaken  of all previous agronomic studies carried  out at Melka Werer so  that full use is made of all existing  information.  Such  a review may assist in  the identification  of the more promising  lines of investigation.
A committee including all interested parties should lx* constituted 
for the purposes of defining the 1990-91 programme of applied research. TAR staff would liave the leading role but would be expected to liaise closely with ADP Consultants and local personnel. As off-station trials are essential if the full range of coruli Lions
in the Stage I area are to be investigated the full co-operation of MAAE, Lhe State Fann Management and the ASP Management must be assured.
FvincLs will have to be made available for the applied research us existing budgets are unlikely to be able to cover additional needs. An initiative may be required from the ADP authorities investigate the allocation of additional funds for reseai-ch purposes.
•12PRELIMINARY REVIEW OF THE IRRIGATION AGRONOMY REPORT BY OTHER ORGANISATION
1. Distribution of the Report and Objectives
Before finalisation  of this report it was considered  essential that the views of the organisations directly  involved  in  the implementation  of proposals should  be considered.  This is particularly  so  because of the number of different organisations involved  and  the desire to  develop  a coordinated  and  integrated approach  to  solving  the problems of crop  production,  water management,  monitoring  and  applied  research  associated  with  this major undertaking.
Accordingly  the preliminary  draft report was sent informally  to  the following  organisations for their consideration  before production  of the final draft report.
a)   Institute of Agricultural Research
b)   Project Control Centre
c)   Ministry of State Farms
Preliminary  comments were received  from the first two  organisations and  their contribution  in  the preparation  of the final draft report is gratefully  acknowledged.  Because of the urgency  in  preparing  the Final Draft Report before the project coordination  meeting  in February  1990  it was felt that the contribution  by  the Ministry  of State Fanns could  usefully  be discussed  at that time and  all points arising  could  be used  to  formulate an  agronomic action  plan  for 1990.
Where possible the contributions given  by  the two  aforementioned organisations have been  incorporated  into  the report,  but for the record they are included as Appendices 1 and 2 hereof.
43APPENDIX 1
I CONSULTANCY SERVICES AGREEMENT FOR SUPERVISION OF CONSTRUCTION - ADPI
by Sir M. Mac Donald & Pantners
comments by A.Kuzniqr
Molkaweror Research Center- L would like to make tha following comments on LIMP reports
l  Irrigation A^ronony Report
o
I agrea with the author on the factors limiting crop production 
levels, particularly on the over irrigation and the low application 
efficiency. But under present conditions it will take a long time
before the efficiency can bo significantly improved. As the fixed
irrigation intervals being followed ^the irrigation scheduling 
suggested on the page 28 may improve water management practices.
However, in addition to the scheduling techniques recommended by tho 
author, I would suggest to use the gypsum blocks for evaluation of tho 
soil moisture status. Obviousley it will not be possible to apply 
this method for all the farm, units. But some areas could be solcctod 
for Buch scheduling method. At prasont the trials are being conducted
at    Molka    Worer    Center    where    irrigation    scheduling    is    controlled    by    tho soil moisture blocks. And results are encouraging.
Moreover, the laboratory facilities aro available at Molka Id such as tho pressure plato and the pressure membrane apparatus. Therefore relationship of readings on gypsum blocks to soil section 
can bo established for difforont soils. Tho soil moisture motors uro inexpensive, US dollars 270.- por unit (fo^ fialifornifcq), tho gypsum block US | 6.- por unit. In addition to the factors limiting product
ion levels such as salinity, lund levelling and ovorirrigution so^us consideration ohould bo given to tho wood control. It Dooms that weeds along channel banks aro u major oourco of wood infestation. And wouds should not bo pormittod to go to sood, they can be controlled by2
periodic burning or with careful use of herbicides.
The author states (page 31) that presently little of the work being attempted, at IAR has relevance for the irrigated state farms© However, 
the mooting had been held An 11 May 1989 where the workplan proposed 
by the writer was discussed# And the representatives of state farms
were also prosent. I believe that suoh trials as evaluation of surface
irrigation, pro-irrigation of cotton and wheat and study of crop water requirements have relevance for the Amibara Project.
In tho part proposals and recommendations page 36 "section 5«3 
the investigation on the field application efficiencies for furrow
irrigation is re commended o Such studies worf already suggested by the writer as the part of the studies on surface irrigation. There were
also such studies at Melka Werer before see A. Kandiah evaluation of furrow irrigation system for cotton. IAR, 1989*
As concerning leaching and land reclamation trials (section 5.5. page 3) the preliminary agreement had been reached between ADPI and IAR that the land to the north-east of the Pilot Drainage Area (Malka Sadi) adjacent to the field AC/lO’B could be used. Tho area had boon selected by Sir. W. Hal crow for their trials on open fiold drains, but the si to 
was finally abandoned. The advantages of that location is discussod in the soil and groundwater Monitoring Programme (ittlP, 1989 page 9).
Tho author found that during July 1989 the frequency of irrigation was generally too high (P.38). However, G. Haider, Melkamu Riga reported over-irrigation during low demrnd period of April and Hay and under irrigation during rest of tho growing season (See ETH /82/014 Field Document No 5 August 1987) • As o on coming tho trials an irrigation froquoncy recommended by tho author (page 38), such experiments had boonAPPENDIX 2
+ ^rc r.
1UL No. "
’"
—Ah a>az Dec, 7 
19 3?
MrlU Wa/cZ
fMftf .lnint’ <rh<J-rtj’f iimiaJ)
(W u-fl-7- flA/* A'
J
T )'> fTtrjfh-'b W>"kn<U’ •7AhA
THE PEOPLE'S DEMOCRATIC REPUBLIC OF ETHIOPIA
WATER RESOURCES DEVELOPMENT AUTHORITY PROJECT' CONTROL CENTRE
Aiiibara Drainage Project
Vater Resources Development Authority
Add.jp Ababa
WATER RESOURCES DEV.AUTnORITY Aniibaru Project Control Center
Sub 1- Commonto on Irrigation /Agronomy Report
July-Auguat 1969 by .Sir H.MacDonald
Vo rofer to your letter Rof No ID 25A/1,dated 14 Nov, 1939 regarding tho above subject.
Au por your roquout, we horoby give uomo comments on few of tho articloo jtatod in tho report.
1, Page 4. Climatic Data;- Under thio article tho variable feature
of tho rocordu of rainfall ip mentioned. Vi th roupuct to thio, 
ovunthough, tho re io no apparent reauonr for the extra-ordinary differences between tho avuilablo rocordu of rainfall,wu would like to give pome
real oituationu on uito.
— Tho ruinfall oiiaractoriutioa of tho region io very erratic,non- uniform A rovoalo a marked Variability with in uhort dictanccu loop than a kilomotor.
- Thuro are 4 other cluue IV ruinguagu uituu with in tho project urea for width uvuilublo rucordu with ruupuct to thut of the LUI give
_____^relatively lower vulUuu A utill varying in magnitude from ouch other.APPENDIX 2
*<■/:■> I: /±-
kcL No. T
••Ah a>dc Dec, 7 IV 50
MdU Wm/ct
~
rM-rfc/ Mot
iiHUth
na>y ihp> oa/^a"]'} rnrKh> u**«u\\&y ••mra
THE PEOPLE’S DEMOCRATIC REPUBLIC OF ETHIOPIA
WATER RESOURCES DEVELOPMENT AUTHORITY PROJECT CONTROL CENTRE
Amibara Drainage Project
Water Resources Development Authority Addio Ababa
WATER RESOURCES DEV.AUTHORITY Amibara Project Control Center
Sub:- Comments on Irrigation /A^ronony Report
July-August 1989 by Sir M.MacDonald
We refer to your letter Ref Ro ID 25A/1,dated 14 Nov,1989 regarding |he above subject.
Ao per your request, we hereby give some comment a on few of the articles jtuted in the report.
1. Page 4. Climatic Data:- Under this article the variable feature
of the records of rainfull io mentioned. With respect to thio, cvunthough, there is no apparent reason:' for the extra-ordinary differunc between the available records of rainfall,we would like to give same real situations on site.
-   The rainfall characteristics of the region is vory erratic,non- uniform A reveals a marked Variability with in short distances loos than a kilometer.
-   There are 4 other cluse IV xuinguugu situs with in the project ar^u for which available records with ruspuct to that of the LUI give relatively lower values A still vurying in magnitude from each other.conducted at Molkaworor 'before and final conclusions were established 
for uso at AlP.^fy^
2. SOIL AlIP GROUNDWATER MONITORING PROORAJ1E FOR 1?S?-1999O-
Pilot Drainage Area Studies/pago 8^
In addition to rehabilitation of equipment and restoring of standard observation procedures by AZP thero is an urgent need <*o
improve hydraulic characteristics of tho Primary Drain SD 2. I* &_
seems that thero is problem with tho slope and outflow of water is difficult. Consequently in the Fields 4^/6 and 4C/7; ° * °- the
u
three test lines (i) 3, D 8 and D 13) only the two drains can be
studies. The outlet of tho drain D 13 is submerged for much of the
time- Similarly in the Fiold 4C/1OA tho drain No D 24 is also submerged 
and the two test lines can bo only invostigatod. Tho improvement of
the Drain SD2 will bo nocessaiy for proposod leaching trials located 
to the north-east of tho Pilot Drainage Area (Helka Sadi). It is
also suggestod that tho IAR activity should also include cooporaticn on monitoring of drains (analysis of results); study of entranco 
resistance and tho relationship between hoad and discharge. The
latter had boon investigated by G.Haidor (FAO export) 3oe. Haider G.
and. Mclkarau Regea 1987. Monitoring of Malka Sadi Pilot Drainage Scheme. Hwevor the IAR activity at Pilot Drainage echetno will include ehlnfjy 
investigation of changos in soil salinity over time. INVESTIGATIONS OF SPECIAL PROBLEMS (Activity 16 - Fig.4)
Moro dotailod. investigation on boron and/or tho volcanio tuff ooila suggostod. for tho IAR cannot bo conducted in ba period 1989-1990. Still tho Molkaweror Contor ia in the prooeos of improving tho laboratory Xacilitioe (instruments and manpowor). UnlesD it is dono (by tho <md of U90) the more complicated, onalysio uro not possible.2
One poosiblo reason for the high rainfall records of the IAR
Btatlon io that the otatlon in surroundod with sufficiently grown 
buohos and treee irorving an wind breaks,but not close to the gunge
an to reduce the guage catoh. The locations of the rest of the rain
gauges are relatively in anopen area. In an area li^e Amlbara where precipitation 1b normally coupled with atrong wind, the absence of
wind shield around the gauges reduces the catch by about 20% imparting 
an up ward acceleration to precipitation about to enter and resultsin 
in a deficient catch.
The variability in the rainfall recorde of the project area hae
been noticed long ago by the Water Management Unit of the P.C.C and 
for practical purposes, we have been using the averages of the available records.
The case can also be taken to the attention of the National Meteoro logical Services Agency for detailed study of the rainfall behavior or
in case there maybe a possibility to reconcile the existing difference
based on the sites'situations.
2.  Page 5. Ground Water Levels and Salinity
In analysing the piezometer data, no account has been given to the existing land use which majorly influences the water level in each of
the piezometer. The 71 paezometers are situated in such away that they give a good coverage of the area. Presently, however, the area these piezometers are representing donot have a uniform and continuous irrigation practice. Some of the piezometers are located in an area where irrigation has been inherently with held because of the throwal of the land out of cultivation for reasons of salinity. State forms have also reinstated 
some of the fields to introduce some salt tolerant crops resuming irrigation after a long period of interruption and this again has its own influence
on the trend of the piezometers readings.
As stated in the report existing GWT records reveal gradual trend towards higher water table levels-contrary to the previous prediction because most of the area where ground water depths wire<2m and the
ils are highly ealiae have arready been out of cultivations and irrigation is also permanently diecontimued. GWT records with in such an area lave evon a falling trend.Thuo, the oet of evento under which the data was collected vary 
from one area to another to the extent that complications arise to 
derive a general conclusion representing the ground water table trend 
of the project area.
Wo suggest a comprehensive GWT record analysis in conjunction 
with the land use details of the project farms to derive better
conclusions.
Page5. Irrigation water supplyUnder this article in trying 
to justify the dependence of the irrigation water supply only on the
river run and the absence of any correlation between irrigation water
use rainfall, the low figure for the 1908 is referred. This appears
to contradict with the statement under article 2. J which mentions
’’exceptional rainfall in August 1900 when 262.4™.• •". Furthermore,
the low figure is because of a considerable hectarage remained unplanted 
to cotton because of shortage of irrigation water during per-iirrigation
4 planting periods of the year.
In addition to the salinity problem which remarkably reduced farm land of the project area, fl actuation in water use from year to year was due
to the fact that,during the last four years,state farms have tried to introduce double crops like broom sorghum,wheat,corn and harricot bean. Water supply records indicated in the report also include irrigation water that has been utilized by these crops.
No mention has, also,been made about total excess irrigation flows rejected back to the river at the extreme north of the project area.
In general,more dotailed approach to the water use would have helped in Indicating that water supply has never depended on the river run, 
except in May and Juno of 1900 when the total river discharge was less than tho cumulative irrigation demand of the project farms.
As far as crop production limiting factors are concerned,the report has ma do a good covorage of the existing problems. If arrangements
r <4
are to be made to effect the proposals and recommendations we hope
CC/
Project Operations and Water Management Department W.R.D.A.
Addis Ababa
*’ _^UVED

< <

Summery

Document: Final Draft Report, July-August 1989

Prepared by: Sir M. MacDonald & Partners Limited, Cambridge, England

Location: Amibara, Ethiopia

Project: Amibara Drainage Project I

Executive Summary

The report examines agricultural production challenges in the Amibara irrigation area, focusing on declining crop yields due to soil salinity, poor water management, and other limiting factors. It provides recommendations for improving production through better irrigation practices, land leveling, drainage implementation, and agronomic improvements.

1. Key Findings

1.1 Production Trends

Cotton yields at Melka Sadi State Farm declined from 38.9 quintals/ha in 1984 to 24.0 quintals/ha in 1988
Banana production has steadily declined since 1975, with yields dropping from 238.4 quintals/ha to just 48.4 quintals/ha in 1988
Significant areas of both crops have been abandoned due to salinity issues

1.2 Soil Salinity

Identified as the most serious factor affecting production levels
Caused by rising groundwater tables due to over-irrigation and poor drainage
Salinity problems are most severe in areas with mixed soil textures
No routine soil salinity monitoring system exists

1.3 Water Management Issues

Over-irrigation is common, especially in banana fields
Irrigation scheduling is rigid and doesn't account for soil moisture conditions
Unlevel fields contribute to poor water distribution and low irrigation efficiency
Average irrigation water salinity is 0.43 mS/cm (C2 classification), but reached C3 levels (high salinity) during peak irrigation in 1988

1.4 Resource Limitations

Aging farm machinery and equipment
Capital constraints for major investments
Need for improved training, especially for tractor drivers and irrigators

2. Major Recommendations

2.1 Immediate Actions (1989-1990)

Implement improved monitoring systems for soil salinity, water quality, and crop production
Conduct detailed topographic surveys to assess land leveling requirements
Improve water management practices and irrigation scheduling
Initiate applied research programs for leaching and land reclamation

2.2 Medium-Term Actions

Implement comprehensive land leveling program
Develop training programs for water management personnel and tractor operators
Establish coordinated research programs between institutions
Investigate alternative crops and cropping systems

2.3 Long-Term Actions

Full implementation of drainage system
Reclamation of abandoned lands through leaching and salt-tolerant crops
Development of sustainable crop rotations, especially for bananas
Policy decisions needed for future of Amibara Settlement Project

3. Key Data Tables

3.1 Cotton Yield Trends at Melka Sadi State Farm

Year	Area (ha)	Total Production (quintals)	Average Yield (quintals/ha)
1984	2633	103,808	39.4
1985	2233	91,000	40.7
1986	2633	103,808	39.4
1987	2633	95,108	36.1
1988	2633	81,671	31.0

3.2 Banana Production Decline

Year	Productive Area (ha)	Total Production (quintals)	Average Yield (quintals/ha)
1975	720	171,641	238.4
1980	700	84,353	120.5
1985	440	51,069	116.1
1988	440	21,307	48.4

4. Institutional Coordination

The report emphasizes the need for better coordination between:

Water Resources Development Authority (WRDA)
Amibara Project Control Centre (PCC)
Middle Awash Agricultural Enterprise (MAAE)
Institute of Agricultural Research (IAR)
Relief and Rehabilitation Commission (RRC)

Conclusion: The report provides a comprehensive analysis of agricultural production constraints in the Amibara irrigation area and outlines a detailed action plan to address soil salinity, water management, and other limiting factors through improved practices, infrastructure development, and institutional coordination.