Vol.4, No.11, 595-607 (2013) Agricultural Sciences
Analysis of determinants of gross margin income
generated through fishing activity to rural
households around Lake Ziway and Langano
in Ethiopia
Dawit Garoma1*, Asefa Admassie2, Gezahegn Ayele3, Fekadu Beyene1
1Collage of Agriculture, School of Agricultural Economics and Agricultural Business, Haramaya University, Dire Dawa, Ethiopia;
*Corresponding Author: dgaromaa@yahoo.com
2IFPRI, Ethiopian Economics Association, Addis Ababa, Ethiopia; aadmassie@yahoo.com
3USAID, Addis Ababa, Ethiopia; ayeleg2002@yahoo.com
Received 16 August 2013; revised 16 September 2013; accepted 15 October 2013
Copyright © 2013 Dawit Garoma et al. This is an open access article distributed under the Creative Commons Attribution License,
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
This article analyzed determinants of gross mar-
gin income from fishing to the rural households
around Lake Ziway and Langano in Ethiopia.
Four districts adjacent to the two lakes were
selected purposively from which 179 respon-
dents drawn randomly. Both primary and sec-
ondary inform ation comprises of househ old struc-
ture and assets, climate factors and supportive
services were organized. Data analysis em-
ployed descriptive statistics, budgetary analysis
and the Ricardian method. Budgetary analysis
showed positive fish gross margin income (GMI)
of ETB 3,023.40 to average fisher. The Ricardian
analysis made use of the climate only model
(Model 1) and comprehensive model (Model 2).
Using Model 1, water level raise due to inflow is
vit al to earn fair income in addition to the rainfall
amount in Season 1, which is supported with
positive impact of precipitation water level in-
teraction on fish income. Impact of precipitation
was positive in Season 3, due to meher rainfall
and withdrawal of fishing labor to join agricul-
ture that minimized over fishing. Using Model 2,
precipitation water level interaction has positive
impact in Season 2 due to better inflow and
Meher rainfall. The interaction term was negative
in Season 4 attributed to decreased water level,
dry weather and less precipitation. The result
also showed positive impact of household mem-
bers participation, participation in the traditional
financial arrangement and capacity to finance
operating costs. The study highlights problems
facing fishing business like: decreased lake size
and water volume, lake turbidity and siltation,
open access and weak institutional arrangement
to guide fishing efforts, wetland farming and ex-
p ansion of irrigation to the lake side, c ost of fi s h -
ing materials, minim um sales price as well as poor
access to the fish market. Hence, strengthening
fishery co-operatives, awareness creation, off-
farm opportu nities, integrated conservation works,
reduced wetland farming and acquaintance to
social networks were suggested.
Keywords: Lake Ziway; Lake Langano; Fishing
Household; Gross Margin; Ricardian Technique;
Fish is highly nutritious, so even small quantities can
improve people’s diets. Fish provides about 20 percent of
animal protein intake in developing countries and this
can reach 90 percent in Small Island Developing States
(SIDS) or coastal areas. Fisheries can also contribute in-
directly to food security by providing revenue for food-
deficient countries to purchase food. Fish exports from
low-income, food-deficient countries are equivalent to 50
percent of the cost of their food imports [1]. The number
of people directly employed in fisheries and aquaculture
is conservatively estimated at 38 million, of which over
90 percent are small-scale fishers. Small-scale fisheries,
and especially inland fisheries, have also often been mar-
ginalized and poorly recognized in terms of contribution
Copyright © 2013 SciRes. OPEN ACCESS
D. Garoma et al. / Agricultural Sciences 4 (2013) 595- 607
to food security and poverty reduction [2]. Capture fish-
ery resources were seen as the quintessential “common
pool” (open to all) resources [3]. Traditional, small-scale
or artisanal fisheries are used to characterize those fish-
eries that were mainly non-mechanized with low level of
production. In spite of the relevance of artisanal fishing
to the economy, only few studies have been carried out to
assess the profitability of the enterprise and constraints
faced by the fishermen [4].
It is against this background that the research sets out
to determine quantitatively the amount of income earned
from fish catch and factors affecting fish income around
Lake Ziway and Lake Langano of Ethiopia. The findings
of the research can assist in identifying the significance
of factors affecting fish income and guide policymakers
and development actors in identifying priority areas of
intervention to improve income from fishing activities to
a given household.
2.1. The Study Area
The study area was four districts around Lake Ziway
and Langano, in Ethiopia. The two lakes are located in
the South-eastern direction of Addis Ababa at a distance
of 175 km and 190 km, respectively. Agro-climatically,
the area classified into midland (1500 - 2500 meter
above sea level) and lowland (500 - 1500 meter above
sea level) with proportion of 30% and 70%, respectively
The two lakes are found in the Great East African Rift
Valley and located between 7˚51ʹN to 8˚57ʹN and 38˚43ʹ
E to 38˚57ʹE. Lake Ziway has a water surface of 440 km2
with an average depth of 2.5 meters. Lake Langano own
241 km2 water surface with an average depth of 17 meter.
The annual fish potential of Lake Ziway and Langano is
estimated to be 2941 tons and 1000 tons, respectively
[20]. Lake Ziway basin covers Dugda, Adami Tulu Jido
Kombolcha and Ziway Dugda districts. Lake Langano
basin covers Adami Tulu Jido Kombolcha and Arsi
Negelle districts. Figure 1 shows the geographic location
of the studied district in Ethiopia and the respective
The study area enjoys bi-modal rainfall. Belg1 rain
usually commences in March & ends in April. Meher2
season takes place from June-August is considered to be
the long rainy season during which major crops like
cereals, pulses, oil crops and the like are cultivated. The
average annual rainfall of the area ranges from 800 mm
to 1100 mm while the mean annual temperature varies
between 11˚C and 29˚C.
The human population of the study districts is 770,799,
of which 22% are in Dugda, 21% in Adami Tulu Jido
Kombolcha, 39% in Arsi Negelle and 18% in the Ziway
Dugda district. With population density of 138 persons
per km2, the average family size was 6.01. The average
land holding is 1.5 hectares per household.
2.2. Data and Analytical Tools
Purposive sampling was used to select the four dis-
tricts. Sample respondents of 179 households were ran-
domly drawn to generate data taking the 2011/2012 pro-
duction year. The sampled households were further de-
composed into 86 fishing households who were fishery
cooperative members and 93 fishing households who
were non-cooperative members.
Primary information captures household general char-
acteristics, fishing participation, fishery cooperative mem-
bership status, asset holding, farmland holding, crops
production, livestock raring, infrastructure services and
Secondary information was obtained from the Na-
tional Meteorological Service Agency (NMSA), Central
Statistical Authority, Federal Ministry, Zonal and district
agricultural offices. Climate data on precipitation and
temperature were obtained from seven Meteorological
stations located around the lakes. Water level data were
obtained from Federal Ministry of Water Resources and
2.3. Method of Data Analysis
2.3.1. Conceptual Framework of Fish Income
A gross margin for an enterprise is its financial output
minus its variable costs [6]. Similarly [7] found positive
net margin per fisher using the gross margin analysis.
Study had using 92 respondents have found income from
fishing to be attractive [8]. By employing descriptive
statistics, budgetary and regression analysis fish farming
is proofed to be economically rewarding and profitable.
The relationship between the endogenous variable and
each of the exogenous variables can be examined using
linear, exponential, logarithm and quadratic functional
forms [9]. To determine the extent of relationship be-
tween socio-economic factors and the level of non-farm
income, four functional regression forms were tried, and
a lead equation was chosen on the basis of R2, F-ratio,
number of significant variables and a–priori expectations
In the study of whether changes in Hartwell Lake’s
water level affect regional economic activity, and by how,
linear and nonlinear regression analysis and other statis-
tical techniques were used. These models were used to
evaluate the strength of the relationships between key
easures of lake-related activity and water levels in
1Belg season represents the shorter rainy season.
2Meher season represents the longer rainy season. m
Copyright © 2013 SciRes. OPEN ACCESS
D. Garoma et al. / Agricultural Sciences 4 (2013) 595- 607
Copyright © 2013 SciRes.
Ziw ay
A bija ta
Shala lake
La k e
A.Tulu J.
Neg ele
East Shewa
We st Ar si
Shala lake
La k e
A.Tulu J.
Zew ay
Ne ge le
1: 1000000
L ocatio n Map of the study di stricts
L ocati on of study dist rcts in Ethiopia
Location of study districts in Zones
Figure 1. Geographical location of the study area.
Hartwell Lake. The inclusion of a quadratic term high-
lights the significance of nonlinear behavior between
lake level and gross sales [11]. To determine the eco-
nomic impact of fresh fish marketing descriptive statis-
tics, linear regression and market margin were employed
[12] that recommend moderate tax rate regulated by
government. Economic analysis of smoked fish market-
ing was conducted using regression model [13]. The re-
sult indicates, 94.6% of the variation in the sales value
explained by household structure, operating and acquisi-
tion costs. Despite the widespread interest, study on im-
pact of climate variable on fish income was limited. In
many cases, studies had on fishing have employed linear
regression and discrete choice models, ignoring the in-
fluence of climate variables in the predictors domain of
the income derived to households [14].
Two main methods have been used in the literature to
study the impact of climate change on agriculture. The
traditional approach is a production function method
which relies on empirical or experimental production
function to predict environmental change and the Ricar-
dian approach that uses economic data on the value of
land to analyze the impact of climate on agriculture.
Linear and quadratic terms are included in the regression
analysis to reflect the non-linearity that are apparent
from field studies. The linear term reflects the marginal
value of climate evaluated at the mean, while the quad-
ratic term shows how that marginal effect will change as
one moves away from the mean. However, this approach
has been criticized for not fully controlling for the impact
of important variables that could also explain the varia-
tion in farm incomes, for assuming that prices are con-
stant and adjustment costless, and for possibly yielding
biased result when land within location is heterogeneous
and land owners behave optimally [15]. Consequently,
Farmers at particular sites take environmental variables
like climate as given and adjust their inputs and outputs
accordingly. The Ricardian method was used to measure
how climate affects current net revenues. This method is
a cross-sectional technique that regresses net revenues on
independent variables [16]. This method has been ap-
plied to selected countries in the low latitudes, namely
Brazil and India, using district level data, and Sri Lanka
and Cameroon [17], using household level data. Origi-
nally presented by Mendelsohn, Nordhaus and Shaw the
Ricardian model is a cross sectional analysis of the im-
pact of climate on land value or farm revenue. In coun-
tries with a large percentage of small farmers and unde-
D. Garoma et al. / Agricultural Sciences 4 (2013) 595- 607
veloped land markets, farm revenue is used. The model
uses a multiple regression approach where the farm
value/land revenue is regressed on climatic variables
such as temperature, rainfall and rate of runoff of rainfall,
geophysical variables such as soil type, soil erosion, sa-
linity, flood probability and wind erosion and economic
variables. The estimated model is then used to predict the
effects of future changes in the climatic and geophysical
variables on farm revenues or land values [18]. By col-
lecting data from farm households in different agro-
ecological zones of the county, net crop revenue per hec-
tare was regressed on climate, household and soil vari-
ables. The results show that these variables have a sig-
nificant impact on the net crop revenue per hectare of
farmers under Ethiopian conditions [19]. The impact of
long term climate on agriculture productivity was exam-
ined by applying Ricardian technique to estimate the
effect of climate change on the smallholder agriculture
sector in Sri Lanka. The study assumes that Net revenue
(NR) depends on climate and a host of exogenous deter-
minants [20]. Study had on impact of climate change on
rice agriculture in Nigeria employed the Ricardian model
[21]. The results clearly demonstrate irrigation as a sig-
nificant technique used by the farmers to adapt to the
climate change. Although research work has been done
on fish management at Lake Ziway [22] little attention
was given to quantitatively determine factors that affect
income derived from fish catch.
Hence, the conceptual framework of this resaerch
work was defined in Figure 2.
2.3.2. Analytical Framework of Fish Income
1) Descriptive Statistics
Descriptive analysis involved the use of frequency
distribution, percentages and tabulation of results. Hence,
results of the survey were summarized using statistical
tools that characterize fishing households.
2) Budgetary Technique
Gross margin (GM) is expressed as:
GMTR TVC (1)
NIGM TFC (2)
where; GM = Gross margin/year; TR = Total revenue
(ETB); TVC = Total variable cost (ETB); NI = Net in-
come (ETB); TFC = Total fixed cost (ETB).
Fixed cost were depreciated using straight line
method represented as:
V = Original value of fixed input; S = Salvage value N
= No. of economically useful live.
Climate variable
Fish Income
Household structure
Household asset
Supportive service
Figure 2. Conceptual framework of the study.
3) The Ricardian Model
The Ricardian method is a cross-sectional approach to
studying agricultural production. It was named after
David Ricardo (1772-1823) because of his original ob-
servation that the value of
land would reflect its net productivity. Farmland net
revenues (V) reflect net productivity. This principle is
captured in the following equation:
,, ,,ii xVPQXFHZGPX
where Pi = market price of cropi; Qi = output of cropi;
X = vector of purchased inputs (other than land); F =
vector of climate variables; H = water flow; Z = vector
of soil variables; G = vector of economic variables such
as market access and Px = vector of input prices.
The farmer is assumed to choose X to maximize net
revenues given the characteristics of the farm and market
prices. The Ricardian model is a reduced form model
that examines how several exogenous variables, F, H, Z
and G, affect net revenues [23]. The standard Ricardian
model relies on a quadratic formulation of climate:
  (5)
where Bo-B5 = Coefficients/parameters; U = Error term.
Both a linear and a quadratic term for temperature and
precipitation are introduced. The quadratic term reflects
the non-linear shape of the net revenue climate response
function (Equation 5). The expected marginal impact of a
single climate variable on farm net revenue evaluated at
the mean is:
1, 2,2ii
EdVdfbBE f i (6)
Linear and quadratic terms are included in the regres-
sion analysis to reflect the non-linearity that are apparent
from field studies. The linear term reflects the marginal
value of climate evaluated at the mean, while the quad-
ratic term shows how that marginal effect will change as
one moves away from the mean. When the quadratic
term is positive, the net revenue function is U-shaped
and when the quadratic term is negative, the function is
hill-shaped. We expect, based on agronomic research and
Copyright © 2013 SciRes. OPEN ACCESS
D. Garoma et al. / Agricultural Sciences 4 (2013) 595- 607 599
previous cross-sectional analyses, that farm value will
have a hill-shaped relationship with temperature. For
each crop, there is a known temperature at which that
crop grows best across the seasons. The relationship of
seasonal climate variables, however, is more complex
and may include a mixture of positive and negative coef-
ficients across seasons.
The change in annual welfare, ΔU, resulting from a
climate change from C0 to C1 can be measured as fol-
 
1UVC VC 0
If the change increases net income it will be beneficial
and if it decreases net income it will beharmful.
3.1. Distribution of Fishing Households
The fishing households were drawn from four districts
found adjacent to Lake Ziway and Lake Langano as
presented in Table 1.
Lake Ziway is accessed by 77.1% of the respondents
found in Dugda, Adami Tulu Jido Kombolcha and Ziway
Dugda districts. Lake Langano is largely accessed by
Arsi Negelle district households (22.9%) in addition to
the 25 households from Adami Tulu Jiddo Kombolcha
district. Majority (90.5%) of the respondents were male
while the female constitute 9.5%. This result is also in
agreement with the traditional gender pattern of fishing
[24]. For details refer Table 2.
As presented in Ta ble 2, Most of (59.2%) the respon-
dents were in the age of 30 - 45 years with minimum
participation of elders (age > 60 years), which accounts
for 1.7%. Consequently, 79.3% of respondents were in-
volved in fishing for more than six years and 95% of
them were in their original place. Only 5% of house-
holds were come to the area because of marriage and
employment opportunity. Respondents were distributed
over a wide range of educational backgrounds with 14%
who did not access formal education (illiterate), 85.5%
Table 1. Fishing household distribution.
Household distribution by District
composition Dugda Adami Tulu Jido
Kombolcha Arsi
Negelle Ziway
Dugda Total
Male 32 52 39 39 162
Female 5 9 2 1 17
Total 37 61 41 40 179
Percentage 20.7 34.1 22.9 22.3 100
Kebele 9 11 10 8 38
Source: Computed from data of 2011/12 household survey. Kebele is the
lowest unit in the government administrative structure
Table 2. Respondents socio-economic characteristics.
Category Frequency Percentage
Age (years)
20 - 29 41 22.9
30 - 45 106 59.2
46 - 60 29 16.2
>60 3 1.7
Original place of respondents
This locality 170 95
Other locality 9 5
Fishing experience (years)
Less than 6 37 20.7
6 - 15 108 60.3
16 - 30 34 19
Education level of respondents
No formal (illiterate) 25 14
Grade 1 - 4 141 78.8
Grade 5 - 8 12 6.7
Grade 9 - 10 1 0.6
Family size
Less than 5.6 135 75.3
5.6 & above 44 24.7
Source of fishing labor
Family labor 175 98
Shared labor 4 2
Ethnic composition
Oromo 160 89.4
Non-Oromo 19 10.6
Religious affiliation
Orthodox 57 31.8
Muslim 103 57.5
Protestant 12 6.7
Wakefata 7 3.9
Farmland holding (ha)
Less than 2.02 117 65
2.02 & above 62 35
Livestock holding (TLU)
None 45 25.1
Between 0 & 4.6 51 28.5
Above 4.6 83 46.4
Copyright © 2013 SciRes. OPEN ACCESS
D. Garoma et al. / Agricultural Sciences 4 (2013) 595- 607
Fishing operating cost
Less than ETB 551 19 11
ETB 551-1185 74 41
ETB 1185-1819 71 40
Greater than ETB 1819 15 8
Cooperative membership status
Member 86 48
Non-member 93 52
Equib financing
No 65 36
Yes 114 64
Access to formal finance
Yes 96 54
No 83 46
Distance from big market
Less than 1 km 79 44.1
1 - 5 km 45 25.1
5 - 10 km 39 21.8
Above 10 km 16 9
Household Adult equivalence Total adult equivalent Percentage
Age less than 10 years 118.1 11.3
Age 10 - 13 years 913.5 86.9
Age greater than 13 years 19.2 1.8
Observations 179 100
Source: Computed from data of 2011/12 household survey.
had primary education, and 0.6% had secondary educa-
tion. The average family size of fishing households was
5.6 with the minimum of 2 and maximum members of 12.
Family labor constitutes the highest proportion (98%) of
labor sources to the fishing households. Consequently,
the total adult equivalence (AE) of fishing households
was 1050.8. The higher proportion of AE (86.9%) was
attached to age group 10 - 13 years followed by below 10
years age (11.3%). The lowest proportion attched to
greater than 13 years indicating less adult members in the
household composition. The majority (89.4%) of the
respondents were Oromo, while 10.9% were from Am-
hara, Hadiya and Guraghe people. The non-Oromo re-
spondents were come to the area in search of employ-
ment and through marriage arrangements since longer
time. The minimum stay of the non-Oromo household
were 12 years, as a result they consider the current loca-
tion as their original place. The majority of them are
Muslim with 57.5%, followed by Orthodox Christianity
of 31.8%. Consequently, 6.7% account for Protestant and
the remaining 3.9% are Wakefata3 followers. For details
refer Table 2.
The average farmland holding is 2.02 hectares to cul-
tivate crops. 65% of the respondents own less than the
average and 35% owned above the average. The mini-
mum and maximum farmland holding is zero and five
hectares, respectively. The available land was wholly
used for crop cultivation during the long rainy season
(June-August). On the other hand, about 42% of the
fishing households were using part of their farmland for
irrigation in the dry season mainly to cultivate vegetables
such as onion, tomato, potatoes, and the like. Only 4.5%
of the fishing households were cultivated their farmland
in the belg season, the short rainy season (February-
April). An average household owns 4.6 TLU of livestock,
while about 25.1% of the respondents did not own live-
stock. With the mean value of ETB 1185 per respondnet,
the minimum and maximum operating expense associ-
ated to fish catch is ETB 258.00 and ETB 5046.00, re-
spectively. For details refer Table 2.
From the study result, 48% of the fishing households
were registered into fishing cooperatives, while 52%
involve in fishing without securing membership status.
Almost 64% of fishers were reported to involve in Ekub4,
a traditional financial arrangements to raise cash to over-
come immediate financial constraints. Despite the poor
outreach and tight criteria, 54% of the respondnets
showed inteerst to use formal financing. Formal sources
are usually accessed for other agricultural and off-farm
activities other than fishing. About 69.2% of the respon-
dents were located within 5 km distance from the nearest
big market. The markets are usually located across all
weather roads and believed to have customers of the fish
products. On the other hand, 21.8% of the respondents
are located at 5 - 10 km and 9% are required to travel
more than 10 km to reach the big market. The results
were presented in Table 2.
The climate variables considered for this study were
precipitation, water level, and temperature. There are
many ways one could represent monthly temperatures
and precipitation data in a Ricardian model. It is not ad-
visable to include every month, because there is a high
correlation between adjacent months, thus three months
average season were used. In this study, climate variables
3Cultural belief among Oromo people.
4Equib is traditional cooperative or traditional self-help group, is a
rotating saving and credit type association whose members make regu-
lar contributions to a revolving loan fund. Its formation is based on
classes of people who have identical (similar) earning or income.
Unlike saving and credit cooperatives, it does not bear interest on the
money saved or collected.
Copyright © 2013 SciRes. OPEN ACCESS
D. Garoma et al. / Agricultural Sciences 4 (2013) 595- 607
Copyright © 2013 SciRes. OPEN ACCESS
were incorporated by classifying the study period in to
four seasons. Accordingly, months were classified as
Season 1 (February-April), Season 2 (May-July), Season
3 (August-October) and Season 4 (November-January).
The seasons were classified considering the seasonal
variability of fishing activities and its market perform-
ance. In certain cases, climate data were not properly
registered. These problem was observed with water level
and temperature data. As a result, some months were left
blank for the climate variables. Hence, the researcher
was dictated to systematically consider the average of
available data alone, ignoring months with blank record.
The average value of months with data record was con-
sidered to obtain the seasonal average data.
From the study result, the minimum and maximum
precipitation record was observed in Season 4 and Sea-
son 3, with mean record of 34.24 mm and 304.88 mm,
respectively. Consequently, the minimum and maximum
temperature record was in Season 4 and Season 2 indi-
cating average of 15.32˚C and 18.77˚C, respectively.
Likewise, the mean minimum and maximum water level
record was 1.01 meter and 1.88 meter in Season 4 and
Season 2, respectively. On the other hand, the interaction
term of precipitation water level indicated mean mini-
mum of 47.82 in Season4 and 447.85 in Season 3. The
deatils of the results were presented in Table 3.
3.2. Fish Income Analysis (Gross Margin
The total fish caught in the study year was 115,825.3
kg. Accordingly, the total revenue from the sales of fish
caught was ETB 984,515.20. The study reveals that the
fishing household has realized an average gross revenue
of ETB 5500.10. The variable cost items comprises of
the expenses of labor, fuel lubricant, repair and mainte-
nance, fish processing sanitation and transportation, food
& drink or entertainment and the like, which is worked
out to be ETB 2476.70 (one USD=17ETB). Thus, gross
margin for each fishermen was calculated as the differ-
ence between the gross revenue and variable costs. Ac-
cordingly, the average gross margin per fishing house-
hold was ETB 3,023.40. Consequently, the net income as
the difference between the gross revenue and total costs
was ETB 1,899.00. The result of the study revealed that
fishing household gets less income from fishing taking
the average family size. This could be partly explained
by the lower price offered in the local market and at the
landing site, and less quantity of fish catch as compared
to the fishery potential of the two lakes due to increased
number of fishermen. For details refer Table 4.
3.3. Factors Affecting Fish Gross Margin
3.3.1. Estimation Issues
For the climate variable we present results of precipi-
tation, temperature in squared term and precipitation
water level interaction due to co-linearity of temperature
and water level in linear terms. Similarly, the correlation
value between precipitation and water level were similar
across the four seasons. Hence, interaction terms of pre-
Table 3. Climate variables.
Variable Obs Mean Std. Dev. Min Max
Precipitation_Season 1 (Feb.-April) 179 149.94 24.45 117.40 174.32
Precipitation_Season 2 (May-July) 179 304.88 45.11 248.94 389.07
Precipitation_Season 3 (Aug.-Oct.) 179 242.06 65.94 174.49 412.30
Precipitation_Season 4 (Nov.-Jan.) 179 34.24 16.02 21.04 63.97
Temperature_Season 1 (Feb.-April) 179 16.79 4.81 11.08 21.27
Temperature _Season 2 (May-July) 179 18.77 3.49 14.39 21.92
Temperature _Season 3 (Aug.-Oct.) 179 17.73 3.69 13.60 21.31
Temperature _Season 4 (Nov.-Jan.) 179 15.32 7.80 6.07 22.79
Water level _Season 1 (Feb.-April) 179 1.03 0.22 0.99 1.77
Water level _Season 2 (May-July) 179 1.88 0.15 1.43 2.15
Water level_Season 3 (Aug.-Oct.) 179 1.48 0.35 0.81 1.13
Water level _Sesaon 4 (Nov.-Jan.) 179 1.01 0.38 0.75 1.20
Precipitation Water level_Season 1 (Feb.-April) 179 159.01 52.49 88.05 209.18
Precipitation Water level _Season 2 (May-July) 179 306.64 55.20 201.64 439.65
Precipitation Water level _Season 3 (Aug.-Oct.) 179 447.85 111.21 249.52 721.35
Precipitation Water level _Season 4 (Nov.-Jan.) 179 47.82 20.42 20.83 113.23
Source: Computed from secondary data, 1981-2011.
D. Garoma et al. / Agricultural Sciences 4 (2013) 595- 607
cipitation and water level were considered in both mod-
els (the climate only model and the comprehensive
model). This simplified the problem associated to co-
linearity of explanatory variables.
In the comprehensive model, we have dropped the
temperature in linear terms across the four seasons due to
co-linearity problem. Thus, the model examines the com-
bined effects of precipitation and water level as climate
factor jointly with household specific characteristics.
3.3.2. The Analysis Result s
1) Dependent Variable: Fish Gross Margi n Income
We carried out an analysis of fish gross margin earned
by households in the study year as a dependent variable
using the result with climate variables only (Model 1)
and the comprehensive model (Model 2) that introduces
household characteristics.
Table 4. Average costs and revenue of fishing household in the
study period.
Items Mean value (ETB)
Fish output revenue
Fish catches (KG) 647.10
Price (ETB/KG) 8.50
Gross revenue 5,500.10
Variable cost
Labor 962.3
Fuel & lubricant 209.1
Motor boat repair & maintenance 461.4
Local boat repair & maintenance 230.1
Gillnet repair & maintenance 247.6
Fish processing sanitation 177.8
Transportation & marketing 188.2
Total variable cost 2,476.70
Gross margin 3,023.40
Fixed cost (Fishing asset depreciation )
Motorized boat* 337.30
Local boat (reed boat) 224.90
Gillnets 188.70
Processing equipments 224.90
Refrigerators & accessories 148.70
Total fixed cost 1,124.5
Net fish income 1,899.00
Source: Computed from data of 2011/12 household survey. *Motorized boat
was donated to the fishing cooperative by NGO (Catholic church, EU
through MOA).
While we do not present the results here, a number of
climate factors and household specific variables such as;
temperature, precipitation, precipitation water level inter-
action, gender, education level, farmland holding, live-
stock holding, irrigation area, whether a household is co-
operative member or not, whether a household uses for-
mal credit source or not and household access to big
market were insignificant and so were dropped as they
were not jointly significant.
The a priori expectation is that the sign should be in-
tuitive for all the explanatory variables and the variables
should significantly accounted for the dependent vari-
ables at 1%, 5% or 10% level.
Based on the analysis result, overall the regression ex-
plains 13% and 65% of the variation in fish gross margin
earned by average household using Model 1 and Model 2,
respectively. Jointly, the coefficients of the model are
significantly different from zero (using F-statics) at 1%
level. However, a great deal of the variation remains un-
measured. This is especially true for the climate variable
only model. There are several sources of possible error,
including the misreporting of fish catch, gross revenue,
operating costs, omitted variables and data quality.
2) Impact of Climate Variable
In the climate only model (Model 1), impacts of pre-
cipitation, temperature and precipitation water level in-
teraction were examined. Whereas, the comprehensive
model (Model 2) considers the impact of climate factors
jointly with household specific characteristics.
Despite the negative impact of precipitation in Sea-
son 1, the interaction term of water level and precipita-
tion has positive impact indicating the importance of
water inflow to enrich lakes and the fish population in
Belg season. The longer fasting period for the Orthodox
Christianity took place in this season during which fish is
heavily consumed in the country. As a result the raise in
catch volume is complemented with better demand in the
season. Consequently, the unit price raises both at the
lake landing site and market places that contribute to
increased income. The lakes receive water through pre-
cipitation, runoff, and inflow from rivers/tributaries.
Complemented with additional precipitation water level
rises that in turn increase fish population. Since the mar-
ginal impact of precipitation water level interaction is
positive in this season, a fishing household is in a posi-
tion to earn extra gross margins of ETB 30,037.00 from
fishing activities in the studied year, which is significant
at the 5% level. The squared term indicates the non-lin-
ear relationship.
In Season 3, the marginal contribution of precipitation
to fish gross margin income is positive and significant at
the 5% level. In Season 3, rainfall is important than run-
off and tributaries. Hence, interaction of precipiation
Copyright © 2013 SciRes. OPEN ACCESS
D. Garoma et al. / Agricultural Sciences 4 (2013) 595- 607 603
water level showed negative impact. During this season,
farmers were occupied with seasonal agricultural activi-
ties. Besides, the demand for fish is fairly stable in the
country. Hence, the number of fishermen reduces due to
the pull effect of farming. Consequently, there is more
water inflow due to runoff and tributaries that adds to
rainfall directly received. Accordingly, household who is
still in the fishing business were in a position to earn
better income. The squared term is a significant showing
non-linear relation.
Due to overfishing of large-bodied fish species, there
has been a gradual shift in fish capture from large and
valuable carnivorous species to smaller, less valuable
species that feed at lower trophic levels [25]. In three out
of nine lakes (L. Tana, L. Ziway and L. Awassa) the fish
community was numerically dominated by small fish
species (i.e. Barbus spp.) which represent a low eco-
nomical value for fisheries.
Similarly, precipitation water level interaction has
positive impact on gross margin income in season 2
(May-July) for the comprehensive model (Model 2). The
longer rainy season took in this season that favor fish
population and create favorable situation to easily collect
the fish. Consequently, the increase in fish catch could
contribute to income earned from fish output sales by
enlarging the sales volume. Compared to the previous
season, the fish population in this period raises signifi-
cantly. In this season, consumers mainly the Christians
are free to consume both animal and fish meat, as there is
no religious obligation to abandon animal meat. Hence,
the sales price is fairly stable in comparison to the pre-
vious months. Despite the stability of fish price, still
fishermen are benefiting from increased volume of fish
catch to derive fair revenue. As a result, the impact of
precipitation and water level interaction on fish gross
margin is positive and statistically significant at 1% level.
The negative sign for the square term in Season 2 reflects
an increase of gross margin income at a decreasing rate.
This could be explained by households engagement in a
seasonal farming activities that demand labor, thus with-
draw labor from fishing and reduce the catch volume at
household level. In this season, fishermen attach priority
to agriculture than fishing due to seasonal nature of cul-
tivation and seeding of crops.
Based on the results of Model 2, precipitation water
level interaction reduces the level of gross margin in-
come derived to average households in Season 4 (No-
vember-January). In this season, precipitation is scanty
and water level reduces. Complemented with the in-
creased uses of the lake water for irrigation, the water
level is believed to be highly depleted compared to the
other three seasons. As a result the amount of fish catch
is significantly reduced that directly affect the gross mar-
gin income of fishing households. Hence, Season 4 is
characterized with poor fish catch and the income de-
rived from the enterprise. Accordingly, the marginal im-
pact of precipitation water level interaction on gross
margin is negative (387), which is statistically signifi-
cant at 1% level. The positive sign of squared term indi-
cates the decrease in fish gross margin could be at in-
creasing rate in Season 4.
The significance of both linear and squared terms in-
dicate the non-linear relationship between climate vari-
able and the fish gross margin. However, depending on
what seasonal precipitation or temperature is being ex-
amined, the marginal impact of a climate variable could
be either positive or negative [26].
3) Impact of Household Specific Variables
Considering the comprehensive model (Model 2)
some of the control variables were significant. Increase
in household members participation in the fishing activi-
ties could increase the gross margin significantly. This is
because, fishing requires energy, sufficient hours to
travel on lakes to catch adequate quantity of fish output.
Hence, the possibility to engage more labor into fishing
could increase the chance of securing more catch that
could be sold. Participation in Equib finance contribution
could promote the likelihood of households to earn more
income from fishing. This financial arrangement is con-
sidered as alternative source of finance among neighbors/
relatives to ease cash constraints. Households who in-
volve in fishing have the chance of getting regular cash
income that could be dropped into equib as a saving
strategy to accumulate cash to finance the fishing busi-
ness and/or other priorities. The cash that could be ac-
cumulated through equib contribution could serve for
bulk purchases of fishing inputs and equipments that in
turn help to expand the business. The marginal impact is
statistically significant at 1% level.
Income generated through fishing is also affected by
the households capacity to finance the required fishing
inputs and associated expenses. Provision of fishing ma-
terials, timely repair and maintenance of the available
inputs, deployment of fishing labor and others are be-
lieved to increase fish income by raising the level of fish
harvest and the gross revenue of fish catch in the studied
year. Thus, financing of additional production inputs to
support the fishing activities could result into increased
gross margin income by 10 ETB, which is statistically
significant at 1% level (Table 5).
3.4. Priority Problems of Fishing Activities
Figure 3 shows the ranking of problems faced by the
fishing households. Decrease in lake size/water volume
was ranked first with 49% response. Impurity of lake
water attributed to turbidity and siltation were ranked in
the second with 31% response. Open access to the lake
Copyright © 2013 SciRes. OPEN ACCESS
D. Garoma et al. / Agricultural Sciences 4 (2013) 595- 607
Copyright © 2013 SciRes. OPEN ACCESS
Table 5. Ricardian regression estimates of the fish gross margin income model.
Variables Model 1* Model 2**
Precipitation Season 1(Feb.-April) 35806.5 (15642)**
Precipitation Season 2 (May-July) 6349.7 (4125.7)
Precipitation Season 3 (Aug.-Oct.) 11139.7 (5469)**
Precipitation Squared Season 1 (Febr.-April) 80.5 (33.3)**
Precipitation Squared Season 2 (May-July) 7.6 (5.6)
Precipitation Squared Season 3 (Aug.-Oct.) 10.6 (5.7)*
Precipitation Squared Season 4 (Nov.-Jan.) 85.7 (69.2)
Temperature Mean Squared Season 1 (Feb.-April) 12.5 (97.5)
Temperature Mean Squared Season 3 (Aug-Oct) 3.4 (185.7)
Temperature Mean Squared Season 4 (Nov-Jany) 7.7(87.2)
Precipitation & Water level interaction Season 1 (Feb-April) 30037 (14687)** 6.6 (148.4)
Precipitation Water level interaction Season 2 (May-July) - 360 (183)**
Precipitation Water level interaction Season 3 (Aug-Oct) 7564.6 (3553.5)** 64.5 (68.4)
Precipitation Water level interaction Season 4 (Nov-Jan.) 15316 (10341) 382 (143)***
Precipitation Water level interaction squared Season 1 (Feb-April) 56.4 (27)** 0.05 (0.47)
Precipitation Water level interaction squared Season 2 (May-July) 4.5 (2.6)* 0.49 (0.28)*
Precipitation Water level interaction squared Season 3 (Aug-Oct.) 5.2 (2.5)** 0.05 (07)
Precipitation & Water level interaction squared Season 4 (Nov.-Jan.) 84.5 (53.5) 2.4 (1.1)**
Gender of fishing household (1/0) 1677 (1562)
Education status of fishing household 159 (280)
Household members involve in fishing 2675 (730)***
Farm land 786 (538)
Tropical livestock unit (TLU) 44.6(97)
Irrigation area 59 (983.2)
Equib finance contribution 0.64 (0.24)***
Fishing input expense (ETB) 9.81 (0.8)***
Fishery association membership (1/0) 2.9 (1069)
Access to formal credit finance (1/0) 122.8 (921)
Distance from big market (km) 49.5 (129.4)
Constant 193814.5 (492870) 40740 (15602)
N 179 179
R2 0.17 0.65
F 1.95 15.6
*Model 1 uses only climate variables as regressors. **Model 2 introduces household specific variables. *Significant at 10% level, **significant at 5% level
***significant at 1% level.
and the fish resource were ranked the third problem with
25% response. This has been the case, due to lose legal
enforcement and absence of clearly defined rules and
regulations in accessing the lake to involve in fishing.
They noted, fishing is considered to be a par time exer-
cise to the people living around lake Ziway and Langano.
Household who has interest to undertake fishing is free
to join the business without obtaining formal permission
from the concerned institution in the area. The use of
ake water for irrigation and expansion of irrigation area l
D. Garoma et al. / Agricultural Sciences 4 (2013) 595- 607 605
Figure 3. Priority problems of fishing activities.
to the lake catchment were ranked fourth with 23% re-
sponse rate. Irrigated land expansion to the lakes com-
mand area was also cited as the newly emerging prac-
tice by households adjacent to the lakes. This problem
has been widespread around Lake Ziway, since it is less
saline with minimum conductivity level as compared to
Lake Langano. High cost of fishing materials was ranked
fifth with 20%. The study noticed, increased costs of
fishing materials and equipments to the fishermen. Ac-
companied with decreased fish catch and the associated
income, the tendency to own fishing equipments with
better capacity were limited among fishermen. Similarly,
the capacity to cover costs of fishing materials exclu-
sively from fish income is oftentimes poor. Also, poor
access to market and minimum sales price were ranked
the six problem facing fishermen with 4% response rate.
This problem were noted by aged household as they sale
the fish catch at landing site, where the price offered is
significantly lower compared to the sales price in the
central market and local urban area (Table 5).
This study is a cross-sectional analysis of the fish in-
come of household, relying on the Ricardian method to
investigate the current impact of climate and household
characteristics on gross margin income. Surveys of 179
rural households were combined with data on precipita-
tion, temperature and water level from a number of
sources. The study found that gross margin income of
fishing households was sensitive to climate especially
precipitation and water level. Considering the results of
comprehensive model (Model 2), the combined effect of
precipitation and water level is positive in Season 2
showing an increase in fish income due to Meher rain
and the consequent raise in water level. In this season,
water level raises due to rain, water inflow through
run-off and tributary rivers with positive outcome on fish
population which in turn increase fish catch and gross
margin income of households. On the other hand, the im-
pact of precipitation and the water level interaction on
fish gross margin were negative in Season 4. The season
is characterized with drier weather condition accompa-
nied with insignificant precipitation and water level.
Thus, they resulted in a decrease in the fish population
and the catch level that considerably reduced the fish
income to the household. The impact of temperature on
fish gross margin was insignificant in all seasons, re-
flecting the importance of precipitation and water level
to enhance the fish population and the catch level. The
result also noted the necessity to have more family
members being involved in fishing, households’ partici-
pation in traditional financial contribution (Equib) to
raise cash and household capacity to finance operating
costs of fishing activities. The marginal contribution of
these variables on fish gross margin was positive sug-
gesting the need to use more of the inputs in the fishing
business. Some of the priority problems facing the fish
population were decreased in lake size and water volume.
Wetland farming and irrigation were common using Lake
Ziway, while Lake Langano is mostly used for sand min-
ing. In addition, lake turbidity and siltation were adverse
to the fish population due to runoff caused by the poor
vegetation cover of the lake catchment. Open access to
the fish population, loose legal enforcement, cost of
fishing materials, sales price, poor access to central mar-
ket and the like were the challenges facing the fishing
community around Lake Ziway and Langano.
Accordingly, based on the study results the following
conclusions were drawn: Fishery cooperatives are struc-
tures supposed to implement community based manage-
ment of the lakes and fish population. They were sup-
posed to create market opportunity exclusive to their
members to enable them benefit from sale of the fish
output at completive price. Despite the fact, currently
their performance is loose which is perhaps limited to
collection of the fish output both from members and
non-members at the lake site. The collected output was
Copyright © 2013 SciRes. OPEN ACCESS
D. Garoma et al. / Agricultural Sciences 4 (2013) 595- 607
delivered to wholesalers, retailers, hotel and restaurants,
and individual customers. Hence, there were no clear
demarcation of services provided to the member house-
hold and the non-members. As a result, fishermen were
reluctant to join fishery cooperatives. Hence, the gov-
ernment and non-government organizations directly
working with the fishing business were advised to clearly
define cooperatives role and responsibility and actively
engage towards its application to proof tangible benefits
accrued to cooperative members as opposed to non-
In the study area, the fish population was exposed to
open access problem. Illegal fishermen are rampant
across the two lakes, and legal enforcements to deter
their action were loose. As a result, overfishing was ob-
served in Season 1 and Season 4, in which farming labor
joins fishing business. This resulted in reduction of gross
margin income obtained by a given household across the
seasons. Hence, massive awareness creation were sought
to change community attitude and perception on limited
resources exploitation. Consequently, identifying and
intensifying other off-farm opportunities is helpful to
generate income to the households and systematically
pull excessive labor from fishing activities.
Due to poor vegetation cover, the magnitude of runoff
was increasing that resulted in lake turbidity and siltation
which affect the fish population and the catch level.
Complemented with the cheap price at the lake site, the
decrease in fish output significantly affects the gross
margin earned by a household. Hence, integrated con-
servation works shall be implemented to minimize impu-
rity of water and to maintain suitable water level to the
fish population.
Wetland farming and irrigation use were common at
Lake Ziway. On the other hand, Lake Langano is used
for sand mining meant for construction purpose. This
practice was a recent phenomenon that adversely affects
the fish population and catch level. Hence, demarcation
of the lake compound to exclusively nurture the fish
population has been required to enhance the benefit ac-
crued to fishermen.
In many instances, provisions to be made through for-
mal structures are limited especially with in the rural
setting. Hence, a better understanding of social networks
is essential, especially by examining how these networks
are established and their working mechanisms. This
could offer more insight and lead to their improvement.
[1] Daw, T., Neil Adger, W., Brown, K. and Badjeck, M.
(2008) Climate change and capture fisheries. University
of East Anglia, Norwich, WorldFish Center, Penang, 2012.
[2] Food and Agriculture Organization (FAO) (2005) In-
creasing the contribution of small-scale fisheries to pov-
erty alleviation and food security. Technical Guidelines
for Responsible Fisheries No. 10, Rome.
[3] Gordon, R.M. (2010) From drain to gain in capture fish-
eries rents. A synthesis study. Food and agriculture or-
ganization of the United Nations. FAO Fisheries and
Aquaculture Technical Paper, Rome.
[4] Adewumi, M.O., Ayinde, O.E., Adenuga, A.H. and Zac-
chaeus, S.T. (2012) The profitability analysis of artisanal
fishing in Asa River of Kwara state, Nigeria. Interna-
tional Journal of Development and Sustainability.
[5] Central Statistical Authority (CSA) (2007) Population
census, Addis Ababa.
[6] Firth, C. (2002) The use of gross and net margins in the
economic analysis of organic farms.
http://orgprints.org/8290/1/ .
[7] Inioni, O.E., and Oyaide, W.J. (2007) Socio-economic
analysis of artisanal fishing in the south agro-ecological
zone of Delta State, Nigeria. Agricultural Tropical ET
Subtropical, 40, 135-149.
[8] Onemolease, E.A. andOriakhi, H.O. (2011) The profit-
ability of artisanal fishing in Asa River of Kwara state,
Nigeria. International Journal of Development and Sus-
tainability. http://isdsnet.com
[9] Olayemi, J. (1998) Elements of Applied Econometrics. A
Publication of the Department of Agricultural Economics,
University of Ibadan, Ibadan.
[10] Ibekwe, U.C., Eze, C., Ohajianya, D., Orebiyi, J.S., On-
yemauwa, C.S. and Korie, O.C. (2010) Determinants of
non farm income among farm households in South East
Nigeria. Journal of Academia Arena, 2, 29-33.
[11] Allen, J.S., Carey, R.T., Dickes, L.A., Saltzman, E.W. and
Allen, C.N. (2010) An economic analysis of low water
levels in Hartwell lake. Clemson University, Clemson.
[12] Ali, E.A., Gaya, H. and Jampada, T.N. (2008) Economic
analysis of fresh fish marketing in maiduguri gamboru
market and kachallari alau dam landing site of Northeast-
ern, Nigeria. Journal of Agricultural Society Science, 4,
23-26. http://www.fspublishers.org
[13] Oluwafunmilola, F.O. (2012) Economic analysis of smoked
fish marketing in Owo local government area, Ondo-state,
Nigeria. Continental Journal of Fisheries and Aquatic
Science, 6, 38-46.
[14] Adewuyi, S.A., Phillip, B.B., Ayinde, I.A. and Akerele, D.
(2010) Analysis of profitability of fish farming in Ogun
State, Nigeria. Journal of Human Ecology, 31, 179-184.
[15] Kurukulasuriya, P. and Rosenthal, S. (2003) Climate
change and agriculture. A review of impacts and adapta-
tions. The World Bank Environment Department.
[16] Mendelsohn, R., Dinar, A. and Williams, L. (2006) The
distributional impact of climate change. Environment and
Development Economics, 11, 159-178.
[17] Molua, E. (2002) Climate variability, vulnerability and
effectiveness of farm-level adaptation options: The chal-
Copyright © 2013 SciRes. OPEN ACCESS
D. Garoma et al. / Agricultural Sciences 4 (2013) 595- 607
Copyright © 2013 SciRes. OPEN ACCESS
lenges and implications for food security in Southwestern
Cameroon. Environment and Development Economics, 7,
529-545. http://dx.doi.org/10.1017/S1355770X02000311
[18] Jain, S. (2007) An empirical economic assessment of
impacts of climate change on agriculture in Zambia. The
World Bank Development Research Group Sustainable
Rural and Urban Development Team.
[19] Temesgen, T. and Hassan, R. (2009) Economic impact of
climate change on crop Production in Ethiopia: Evidence
from Cross-section Measures. Centre for Environmental
economics and Policy in Africa (CEEPA). Journal of Af-
rican Economies, 18, 529-554.
[20] Mohamed, A. and Kurukulasuriya, P. (2006) Application
of the ricardian technique to estimate the impact of cli-
mate change on smallholder farming in Sri lanka.
[21] Joshua, A., Abiodun, A. and Hassan, R. (2010) Impacts of
climate change on rice agriculture in Nigeria. The Joint
3rd African Association of Agricultural Economists
(AAAE) and 48th Agricultural Economists Association of
South Africa (AEASA) Conference, Cape Town, 19-23
September 2010.
[22] Felegeselam, Y. (2003) Management of Lake Ziway fish-
eries in Ethiopia. Master’s Thesis.
[23] Mendelsohn, R., Nordhaus, W.D. and Shaw, D. (1994)
The impact of global warming on Agriculture: A Ricar-
dian Analysis. The American Economic Review, 84, 753-
771. www.jstor.org
[24] Williams, S.B. and Awoyomi, B. (1998) Fish as a prime
mover in the Economic life of Women in a fishing Com-
munity. Proceedings of the 9th International conference
of the International Institute of Fisheries Economics and
Trade, Tromso, 8-11 July 1998.
[25] Williams, S.B. and Awoyomi, B. (1998) Fish as a prime
mover in the economic life of women in a fishing com-
munity. Proceedings of the 9th International conference
of the International Institute of Fisheries Economics and
Trade, Tromso, 8-11 July 1998.
[26] Kurukulasuriya, P. and Mendelsohn, R. (2008) A Ricar-
dian analysis of the impact of climate change on African
cropland. African Journal of Agricultural Research, 2,