Irrigation and agriculture development in Africa: Impact on water quality and ecosystem health in the Ethiopian

highlands

Seifu A Tilahun et al., 2019Bahir Dar Institute of Technology

Bahir Dar University

15 July 2019

Study aim

• Purpose: Evaluate the effect of intensification on water quality: shallow groundwater, streams, lake water bodies & biomass

• Location: Amhara region, Ethiopia

Different intensification levels & topographyRobit (intensified)

~1000 haDangishta (less intensified)~5700 ha

• Agricultural land use: 80% (Robit) and 60% (Dangila)• Slope class: 0-43% (Robit) and 0-28% (Dangila)• Irrigation in Robit > Dangila• endosulfan (α+β) 256ml/ha

90Kg/ha DAP & 46Kg/ha UREA

72 Kg/ha DAP & 43Kg/ha Urea

90 Kg/ha DAP & 46 Kg/ha Urea0.4 m3 Organic fertilizer

Watershed land use pH OM(%) TN(%) Avl.P(mg/l)Robit Bata Farm land 6.0±.18 1.9±.8 0.09±.04 7.7±1.9

Grazing land 5.7±.18 3.8±.82 0.188±.04 4.7±1ForestKhat irrig.

5.6±.365.8±.13

3.5±1.83.2±1

0.17±.090.16±.05

5.4±1.79.2±4.4

Dangishta Farm land 5.1±.18 3.2±.6 0.16±.03 15.8±4Grazing land 4.8±.15 4.5±.5 0.23±.02 9.5±2Forest 5.1±.05 3.6±.5 0.18±.02 11.6±1.7

Mean values ±SD of soil chemical parameters from different land uses

Soil characteristics

• Soil pH values were higher in Robit Bata than Dangishta

• Soil OM , TN and available P (Robit < Dangila)

Typical landscape

Groundwater Level Variation Overtime

Groundwater table variations vary across the two watersheds due to

rainfall distribution and irrigation level

Dangishta

Year

1/201

4

5/201

4

9/201

4

1/201

5

5/201

5

9/201

5

1/201

6

5/201

6

9/201

6

1/201

7

5/201

7

9/201

7

1/201

8

5/201

8

9/201

8

1/201

9

Gro

undw

ater

dep

th(m

)

0

2

4

6

8

10

12

14

16

18

20 0102030405060708090100110120130140150

UpslopeMid-slopeValley bottomRainfall

Robit Bata

Year

1/201

4

5/201

4

9/201

4

1/201

5

5/201

5

9/201

5

1/201

6

5/201

6

9/201

6

1/201

7

5/201

7

9/201

7

1/201

8

5/201

8

9/201

8

1/201

9

0

2

4

6

8

10

12

14

16

18

20

Dai

ly R

ainf

all(m

m)

0102030405060708090100110120130140150

UpslopeMid-slopeValley bottomRainfall

Groundwater Nitrate

• Higher level NO3- was

observed during rainy

season

• Irrigation months (Oct-Dec)

higher NO3- compared to dry

non-cropping period (Jan-

May)

• NO3- signficantly lower in

valley bottom and lowest in

Dangila (P<0.001)

• Rainy season: risk of levels

above EPA drinking

guidelines (10 mg l-1)

Groundwater Phosphorus

• Rainfed cropping season

PO43- was significantly

high in both watersheds (

P=.018)

• Higher concentration

levels in Robit likely

related to relase of PO43-

from soil in the landscape

because of low soil OM

(Robit < Dangila in OM)

P-PO

43- (m

g.L-1

)

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35Mid-slope Mid slope

P-PO

43-(m

g.L-1

)

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35Upslope Upslope

Dec/

16

Jan/

17

Feb/

17

Mar

/17

Ap

r/17

M

ay/1

7

Jun/

17

Jul/1

7

Aug/

17

Sep/

17

Oct

/17

No

v/17

De

c/17

Ja

n/18

Fe

b/18

M

ar/1

8

Apr/1

8

May

/18

Ju

n/18

Ju

l/18

Au

g/18

Se

p/18

O

ct/1

8

Nov/

18

Dec/

18

Jan/

19

P-PO

43-(m

g.L-1

)

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35Valley bottom

Dec/

16

Jan/

17

Feb/

17

Mar

/17

Ap

r/17

M

ay/1

7

Jun/

17

Jul/1

7

Aug/

17

Sep/

17

Oct

/17

No

v/17

De

c/17

Ja

n/18

Fe

b/18

M

ar/1

8

Apr/1

8

May

/18

Ju

n/18

Ju

l/18

Au

g/18

Se

p/18

O

ct/1

8

Nov/

18

Dec/

18

Jan/

19

Valley bottom

Processes control contaminants in GW

• MLR analysis produced statistically

significant models( P<05)

• Cloride and ΔH positively contribute to

PO43- in both watersheds

• Chloride, potassium, and fluctuation in

groundwater table (ΔH) positive contribute

to N-NO3- due to dissimilation and

dinitrification

N-NO3-=.5+ (.2Cl-) + (-.317Cl-/N-NO3

-) + (.156K+) + (.232ΔH)

P-PO43- = .027+ (-.006Cl-) + (.008ΔH)

ΔH=monthly groundwater table level fluctuation (m)

0

2

4

6

8

10

0

2

4

6

8

10N

-NO

3-Streams: Nitrate and Ammonium

Both nitrate and ammonia were dominant in stream during rainy period Only nitrate was high during dry period.

0

2

4

6

8

10

Nov

-16

Dec-

16

Jan-

17

Feb-

17

Mar

-17

Apr-

17

May

-17

Jun-

17

Jul-1

7

Aug-

17

Sep-

17

Oct

-17

Nov

-17

N-N

H 4+

0123456789

10

Nov

-16

Dec-

16

Jan-

17

Feb-

17M

ar-1

7

Apr-

17M

ay-1

7

Jun-

17Ju

l-17

Aug-

17

Sep-

17O

ct-1

7

Nov

-17

Dec-

17

Jan-

18

Robit BataDangishta

00.20.40.60.8

Nov

-16

Jan-

17

Feb-

17

Apr-

17

May

-17

Jul-1

7

Sep-

17

P-PO

43-

0

0.2

0.4

0.6

0.8

Nov

-16

Dec-

16

Feb-

17

Apr-

17

May

-17

Jul-1

7

Sep-

17

Oct

-17

Streams : PhosphorusDangila

Dissolved P was high during rainy period in high-flow in the order of 0.6 mgP-PO4

3-.L-1

Robit Bata

Total Nitrogen: spatial

12

August December March

• Concentration of TN is generally lower than WHO permissible limit• Decrease in load and increase in biological activity drive lower TN

concentration.

Ateka et al. (draft)

Total Phosphorus: spatial

13

August December March

• Largest freshwater body (Lake Tana) started to exceed the 0.2 mg/L of P -minimum level for eutrophication

• Increasing trend in TP from Aug 2016 to Mar 2017 could be with high internal loading

Ateka et al. (draft)

Pesticides in groundwater

• Endosulfa-α concentrations were high in rainy period (August) and irrigation (October) period. The values exceeded MAL- value of EU 0.1 µg L-1

• Endosulfa-β in groundwater was not detected during study period

Endosulfa-β

May-16

Aug.-16

Oct.-16

Mar.-17

0.00

0.05

0.10

0.15

0.20

Sampling period

Con

.(µg.

L-1)

Pesticides in Stream

• Endosulfa-α concentrations were high during rainy period (August) and irrigation (October) period. The values exceeded MAL- value of EU 0.1 µg L-1

• Endosulfa-β and 2,4-D in streamflow were not exceeded EU MAL in study period

Residue levels in onion and khat biomass

B) isomer

Raw-onion

Boiled-onion

Khat0

2

4

6

8

10

12

14

A) isomer

Raw-onion

Boiled-onion

Khat

Res

idue

( g.

kg-1

)

0

2

4

6

8

10

12

14

The EU pesticides regulation for α-and β-endosulfanresidue in bulbs and vegetables recommended a tolerance level of 100 µg.kg-1

Key messages for sustainable development

• ILSSI project provided evidence for critical institutional

changes, guidelines and monitoring mechanisms to regulate

agrochemical use and occurance in water bodies

• Increase smallholder awarness on the impact of agrochemicals

on water quality and human health needed

• Target and promote intensified SSI in suitable areas through

evidence-base on agriculture-water-environment-health system

• Reverse degradation and rehabilitate watersheds (river basins)

for improved SSI

• Link watershed management with irrigation

Thank youThis research was supported through Feed the Future Innovation Labs. We are grateful to Sustainably Intensified Production Systems Impact on Nutrition (SIPSIN-IWMI) project and Bahir Dar Institute of Technology for funding the field research. The Innovation Lab for Small Scale Irrigation provided support for institutional engagement and supporting materials.