<<

seed Production and Poverty Reduction: Case of

EVelyne ALazarol and Daina M. Muywanga2

lDepartment of Agricultural Economics and Agribusiness, Sokoine University of Agriculture, P.O.Box 3007,

2District Agricultural and Livestock Development Department, Dodoma Rural District

Abstract

The contribution oj seed production to poverty reduction at household level in the participating villages was studied. Food security and household income was used as proxy indicators oj poverty. Data was collected through a cross section survey using a structured questionnaire. The average incomes / Adjusted Adult Equivalent Units /year were TAS 77, 768/= and TAS 70, 303/= Jor quality declared seed (QOS) producers and non-producers, respectively. A t-test analysis showed that the diJference in income was not statistically significant (p>0.05). The average income Jor QOS users and non-users was TAS 76, 199/= and TAS 67, 168/=, respectively. The diJference in income was also not statistically SignifIcant. These results are due to the Jact that 2002/03 season in which the study was conducted was a bad year, resulting Jrom a general rainJallJailure in the area. Results oj a regression analysis indicated that use oj QOS )

2 contributed Significantly in explaining the variation in income per Adjusted Adult 1

0 Equivalent Units in the study area. 2

d Food security analysis shows that 37% and 28% oj the sampled QOS users and QOS e t a non-users respectively indicated that their households were Jood secure. The paper d ( concludes that QOS contributed to householdJood security and income. However, the r e

h extent oj impact oj QOS is to a large extent influenced by the rainfall. s i l Introduction oj rainwater harvesting technologies is recommended if Jarmers are to b u attain the intended benefits Jrom QOS. P e h t

y Key words: Seed Production, Food security, Income, Poverty b d e t n

a Introduction r g URT,2003). e c n t has been estimated that in It is recognized that future e c i l I , about 42% of agricultural progress depends on r e

d households regularly have the intensification of production n u inadequate food (URT, 2001b). Any rather than the expansion of y a strategy to address rural poverty cultivation (Muliokela, 1997). w e t and food insecurity for the Hence the use of quality seeds, a

G majority of the rural population in along with other inputs and t e Tanzania must involve actions to appropriate cultural practices, is n i b improve agricultural production an-effective way of increasing crop a S

y and farm incomes (URT, 2001 b; production and productivity. b d e *Corresponding author Tanzania J.Agric.Sc. (2007) Vol.8 No.2, 161-172 c u Accepted March, 2008 d o r p e R 162 Lazaro, E.A. and Muywanga, D.M.

Seeds carries genetic potential of household incomes and food the productivity of other inputs, security. Quality declared seed the main role of the other inputs production system has impact on in crop production is to exploit to both producers and users of a maximum the genetic potential quality declared seed. The of the seed. Thus, seed quality is a producers are expected to benefit prerequisite for improved through sales of the seed that they agricultural productivity. produce. The benefits to seed Empirical evidence shows that users are to be attained through most of the seeds used in higher levels of crop productivity Tanzania are unimproved seeds resulting from the use of improved (Mbwele et at, 2000). The formal seed (quality declared seed). The seed sector has been able to objective of this paper therefore is supply only 4% of the seed sown to assess the contribution of on by farmers in Tanzania. The farm seed production to household remaining 96% originates from the income and food security of QDS informal sector such as~ farm­ prod ucers and users in selected saved seed and farmer-to-farmer villages in Dodoma Rural District. seed exchange (Mbwele et at, 2000; Temu and Mtenga, 2000). Materials and Methods On-farm seed production was The study area implemented in Dodoma between

) The research work presented in 2

1 1998 and 2002 as an alternative this paper was conducted in six 0 2 approach to supply relatively villages namely: Mvumi-Makulu, d e t better quality seed than farmer Chibelela, Mwitikira, Mundemu, a d

( saved seed. The initial seed for Bahi -Sokoni and Chalinze in r e seed production is of known h Dodoma Rural District in Dodoma s i l pedigree (approved progeny) to Region. According to URT (2002) b u

P maintain genetic identity and is mostly semi-arid e h purity. The seed that is produced due to low and erratic rainfall. t

y is identified as Quality Declared b About 85% of estimated

d Seed (QDS). QDS is defined as e population (2002) in the region t n seed produced by a registered seed a lives in rural areas relying on r g producer that conforms to the agriculture and livestock keeping e c n minimum standards for the crop activities for their subsistence and e c i l species concerned and which has income. Maize and sorghum are r e been subjected to the quality the most important in terms of d n control measures outlined in the u hectares and production. In y (URT, a seed production gUidelines 2000/01 season, maize and w e 2001a).

t sorghum accounted for 42% and a One of the strategies of the G 26% respectively of total hectares t e programme is to operate the on under food crops in the region n i b farm seed production process on followed by pearl millet that a S commercial basis, so that those accounted for 15% of total y b who participate will realize the hectares under food crops in the d e c benefits (URT, 1994). The benefits region. u d

o as measured in terms of improved r p e R Seed Production and Proverty Reduction 163

Data collection bl ... b5 = Regression coeffici ents for Data were collected in two stages: the independent variables informal survey and forma l survey. XI = Adjusted household size During informal survey focus X2 = Farm size (area cultivated) X3 = Income from n activities group discussions (FGO) were ~ = Value of crops produced per carried out in all the six villages household using a checklist of questions. Xs =Source of seed (or otherwise) Formal survey was carried out using a structured questionnaire E = Error term representing a in four (Mvumi-Makulu, Chibelela, proportion of the variance in Mwitikira and Mundemu) out of income of the household that was the six selected villages. A sample unexplained by the regression of 160 farming households was equation. randomly selected and interviewed. On-farm seed The tables for: Adult Equivalence producers were in small number Scales: Index of Calorific so they were purposefully selected. Requirements by Age and Sex for East and Index of household economies of scale were used in Data analysis determining the adjusted adult Cross tabulation was used to equivalent units (AAEU) (Collier et summarize and present qualitative aI., 1990). Household AAEU,

) average crop value and crop area 2 data . T-test and chi-square were 1

0 cultivated were transformed into

2 used for significance test. Linear

d natural logarithm in order to fit in e multiple regression was used to t

a the model. d determine the contribution of seed ( r e to household income. Household

h Results and discussion s i l income was used as a dependent Seed production b u variable (as an indicator for The results show that 20% of P

e income poverty). Independent interviewed households were seed h t

y variables included: adjusted producers under the quality b declared seed programme while d household size, farm size (area e t the rest (80%, n=160) were non­ n under cultivation), non-farm a r producers. The seed producers

g activities, value of crops and

e were involved in seed production c source of seed (on-farm or n

e over different time periods. The c otherwise). i l

time ranges from one to five years. r e

d Table 1 shows that 53% of seed

n The variables used to fit the model

u producers have participated in

y were as shown below:

a seed production for 3-5 years. This w The model: Y = a+b 1xl+b2 X2+ ...... e implies that the study villages had t a b n Xn + E a long experience in QOS G t

e production. It is therefore expected n i Where: that the awareness of the concept b a

S Y Income per adjusted adult and knowledge of QOS among the

= y

b communities in the study villages

equivalent units (AAEU) d e a Intercept of the equation is high. c = u d o r p e R 164 Lazaro. E.A. and Muywanga. n.M.

Table 1: ResEondents EarticiEation in seed Eroduction Number of Number of res20ndents by village - years Mvumi Chibelela Mwitikira Mundemu Total - Makulu (n=13j (n=5j (n=12j (n=34j (n=4j n % n % n % n % n % 1 4 31 2 40 1 8 7 21 2 1 25 2 15 1 20 5 42 9 26 3 3 75 2 15 2 40 4 33 11 32 4 2 16 1 9 3 9 5 3 23 1 8 4 12 Total 4 100 13 100 5 100 12 100 34 100

The crops that were involved is important in the study area for included cereal crops (sorghum food security. This is because the and Pearl millet). legumes (pigeon area experiences a semi-arid type peas and groundnutsL ~il seed of climate. with very unreliable crop (sunflower) and a variety of rainfall pattern thus sorghum vegetables. The results show that being drought tolerant is 69% of seed producers produced important in mitigating crop loses sorghum seeds (Table 2). Sorghum due to rainfall irregularities. ) 2

1 Table 2: Distribution of households by type of croE in seed Eroduction 0 2

Crops Respondents per village d e t Mvumi Chibelela Mwitikira Mundemu Total a d

( Makulu (n= 13) (n=5) (n= 10) (n=32) r e (n=4) h s i l n % n % n % n % n % b u 11 P Sorghum 3 75 85 4 80 4 40 22 69 e h Pearl millet 2 17 2 6 t

y Sunflower 1 25 4 31 5 15 b

d Pigeon peas 1 25 4 31 e 6 18 t n Groundnut 3 7 3 9 a r g

Vegetables 4 10 4 12 e c Figures do not add to 100% due to multiple responses n e c i l r e Use of quality declared seeds regard to seed production was d n high. However. only 51 % of the u Before analyzing the extent of use y

a of gos it was important to assess sample households had used on w e the awareness of the on farm seed farm seeds at least in one season t a from 2001/03 (Table 3). The G production process. Data t e presented in Table 3 show that analysis of the methods by which n i b 89% of respondents indicated that farmers accessed gos shows that a S they knew at least one of the seed 39% and 11 % of respondents that y b producers in the village. Hence. used gDS obtained through cash d e c awareness among farmers with purchasing and purchases on u d o r p e R Seed Production and Proverty Reduction 165

credit respectively (Table 3). The fulfills the aim of introducing willingness of farmers to buy QDS seeds production in rural a reas so is an indication that there is a that fanners can access improved demand for quality seeds among seeds timely and at affordable the communities. Hence, this price.

Table 3: Respondents awareness, use and access to gns (n = 160) Variable Respondents n % Awareness KnoWs seeds producers 142 89 Do not know 18 1 1 Total 160 100 Use of gns Used 82 51 Never used 78 49 Total 160 100 Access to gns On credit 9 11 By cash 32 39 As gift 18 22 Exchange with grain 15 18

) Farm saved 8 10 2 1

0 Total 82 100 2 d e t a d (

r Figure 1 shows the trends in there was any contribution to the e h s percentages of seed users that farmers' living standards. The i l b were increasing from 5% in increasing proportion of seed u P 1997/98 to 33% in 2000/03. The users is an indication of the e h t

results show that farmers have acceptability of the seed y b used seeds more than five years; technology among farmers. d e t this justifies the assessment if n a r g e c n e c i l r e d n u y a w e t a G t e n i b a S y b d e c u d o r p e R 166 Lazaro. E.A. and Muywanga. D.M.

50 ~ ~ 40 o Jvl/maku1u r-- e- • ChibeleJa . :: 30 r- c II r- f- OI';! v, .. i ttlrjra j 20 :1. " r- I . ~ o Ivh.llldernu ~ t ICi • Total c.. rII[ l- o rLlr I I 1997/98 1998/99 1999/0[1 2000/01 2001/02 2002/03 Sea.!lDn

Figure 1: Proportion of gns users by village over time

Contribution of seed to producers. The average incomes poverty reduction -, /AAEU/year were TAS 77, 768/= An assessment of poverty reduction and TAS 70, 303/= for QDS effects of seed production was done producers and non-producers, at two levels. First level. the respectively (Table 4). The results assessment is done at the level of of t-test showed that the difference income between QDS producers

) QDS producers and second at the 2

1 level of QDS users. Household and non-producers was not 0 2 income and food security are used statistically significant (p>0.05). d e t This can be explained by the fact

a as poverty indicators for QDS d

( that QDS production is done at users. However for QDS producers r e small scale that it may not be h only household income is used s i l expected to generate a substantial b because poverty reducing pathway u proportion of the total household P

of QDS production is direct e

h income. Under the QDS

t through income generated from

y production, each farmer is allowed b sale of seed.

d to cultivate only 0.4ha of seed. e t n According to the QDS production a On farm seed production and r g household income regula tions. farmers can only e c produce quantities that can be sold n Adjusted consumption expenditure e c i within the same community (URf, l of the households was used to r e compare means of income per 2001 a) and therefore this d n emphasizes small scale rather than u adult equivalent per year between

y large-scale seed production. a QDS producers and non- w e t a G t e n i b a S y b d e c u d o r p e R Seed Production and Proverty Reduction 167

Table 4: Comparison of income per AAEU per year (n = 160) Variables No of Mean Std t-value p-value cases Deviation Seed 32 77768 38964 .994 . 163ns producers Non- 128 70303 33939 producers Ns=Not statistically significant at p>O.05

Quality Declared Seed use difference in income between gDS and household income users and non-users was not very Adjusted consumption expenditure large. These results can be due to of the households was used to the fact that the 2002/ 03 season in compare means of income per which the study was conducted adult equivalent per year between was classified as bad year. During gDS users and non-users. The the season most households were adjusted average incomes for gDS facing the problem of tranSitory users and non-users were TAS 76, food insecurity that was a result of 199/= and TAS 67, 168/=, a general rainfall fail ure in the respectively (Table 5). The area.

) Table 5: Comparison of income per AAEU per year (n = 160) 2 1

0 Variables No of Mean Std Deviation t-value p-value 2

d cases e t a Seed users 82 76 198.6 35870.39 1.643 .051ns d (

r Non-users 78 67 167.7 33670.53 e h Ns=Not statistically significant at p

contribution of seed use to area cultivated were statistically d e t household income / AAEU. significant in explaining the n a Households' consumption r variation in income per AAEU in g

e expenditure was used as an the study area. All the variables c n indicator for household income. e tested indicated positive influence c i l The model included all except for adjusted household size. r e d variables that were predetermined which had negative influence. The n u to affect the households' implication is that the use of gDS y a consumption expenditure. The contlibutes positively to household w e t following factors were retained: Use income. The attribution is that the a G of seeds; Household AAEU; Average use of improved seeds by farmers t e n crop value; Crop area cultivated improves crop productivity in tum i b a and Participation in income this leads to higher levels of crop S

y generating activities (IGAs). sales (assuming all other factors b

d The results of the regression constant). The results are e c u analysis are presented in Table 6. comparable with other Similar d o r p e R 168 Lazaro. E.A. and Muywanga. D.M.

studies done in Tanzania. For adult equivalent meaning lhat only example Tanzania Agricultural few or the adults earn enough while Research Project II (TAR? lI-SUA others are dependants. 2002) reported that there was The model results reflected increase in farmers' income due Lo F-statistics of 11.109 significant at use o[ improved varieties or potato, 0.01. At 0.01 levels of significance rice and cassava in project the proposed equation is participating villages. Household acceptable to explain the size (adult units in the household) relationship between the household was negatively correlated with income and use of seeds, adjusted income per ad ult equivalen t and household size. average crop value. was statistically significcmt (0.01). crop area cultivated and This is because the higher the participation in income generating household size (adult equivalent activities variables. units) the lower the income per

Table 6: Regression results on factors influencing household income /AAEU Predictor Adjusted Beta (b "' J Coefficient T-value P-value (Constant) 22.744 0.000 Use o[ gos 0.143 0.068 1.989 0.048" Household AAEU -0.500 ) 0.081 -6.540 0.000*" 2

1 Mean crop value 0.178 0 0.044 2.113 0.036* 2 Crop area 0.165 d 0.064 1.989 0.048* e t Participation lGAs 0.092 a 007 1.301 0.195ns d

( F -statistics= 1 1. 109; *=Signi ficant at 0.05; " =Significant at r 0.01; ns = Not e significant h s i l b u Quality Declared Seed use P g DS non-users respectively e and householdfood h security illdicated t that their households

y Household food security is an \-V '-T e b food secure. The

findings d important livelihood indicator e [or inJicaLe that a large t number o[ n rural households. Food secUlity food a secure households were [rom r g directly dependents on what a seed users. This e means that the c household can produce. n Seed is use o[ seeds leads to increased e c one i of the inputs that l can crop production

at the household r cont e ribute to an increase in level. Further analysis d on food n quantity o[ food produced on farm. u security was done using the y Respondents were requested a to adjusted adult equivalent units w carry out their e own assessment o[ (URT. 1999). t Table 8 indicates that a household food security. The G the c1Vcrage cereals production per t respondents' e responses presented AAEU wa ~ 278 kg per person n per i in Table 7 b show thaL 37% and 28% year among gos users and a 204 kg

S o[ the sampled gos users and among gos non-users. y b d e c u d o r p e R Seed Production and Proverty Reduction 169

Table 7: Propor tion of food secure households by AAEU -Food-- - security QDS 'users gDS on users Total status N % 0/0 N % Food secure 30 37 22 28 52 33 Food insecure 52 63 56 72 108 67 Total 82 100 78 100 160 100

These data show that, on average in average AAEU of cereal between seed users were more foo d secure households using QDS and those compa red to non-users. T-test using non-QDS was statistically analysis showed that the difference significant (p<0.05)-

Table 8: Av erage AAEU of cereal for seed u s e r s and non-users Compared No of Mean Standard t-value p-value [J et irs cases Deviation - QDS users 82 278 231.4 2.409 0.016** g DS Non- users 78 204 149.3 ]\i13: '<=SldtiSlicalIy si.Q,nilka nl at p

( lower proportions of the non-poor The rural poverty line established r e include: 1) the reference year h during 2000/01 Household Budget s i l Survey was used in the h ousehold (2002/2003) was categorized as a b u bad year due to shortage of rainfall P poverty analysis. The poverty line e h 'c s extrapolated and inflated to which resulted into lower crop t

y harvest. 2) It is likely that the

b '1'A 87,799/= per person per year.

d harvested crops were used for e The poverty a nalysis shows that t n 2GI)lu and 18% of QDS users and h ome consumption and there was a r g non-users were non-poor no s urplus crop harvests for sale. e c Therefore, a low consumption n n 's pectively (Table 9). It was e c i expenditure was recorded that l expected that a higher proportion r e oi QDS users would be under the made the majority of households to d n fall below the poverty line.

u nCJIl -poor category resulting from y a w e t a - rable 9 : Households poverty status (n = 160) G t f U\'erty Seed u sers Non-users Total e n i ";f atus 0;0 % % b No of No of No of a S

cases casf'S cases y b Non poor 2 1 26 14 18 35 22 d e c Poor 6 1 74 64 82 125 78 u d o Total 82 100 78 100 160 100 r p e R 170 Lazaro. E.A. and Muywanga. D.M.

Conclusions and seed demand in communities. recommendations development of distribution channels to ensure efficient supply Conclusions of improved seed in rural areas and The presentation in this paper is building human capacity in based on a study conducted in a management of genetic resources production system that is largely within the communities. rain fed. Thus the effect of the Introduction of rainwater rainfall pattern and distribution harvesting technologies is also very cannot be overlooked in this important if fanners are to attain analysis. Failure of rainfall the intended benefits from QDS. drastically reduced the expected The emphasis of rainwater economic gains from introduction harvesting technologies reqUires of improved seeds. Failure of technical support of agricultural rainfall does not only red uce the extension services from both expected returns to the introduced government and non-government seeds but also results into loss of organiza tions. improved varieties. The use of a pre-determined poverty line. to assess impact of Acknowledgement gDS use on household poverty The authors acknowledge the funding from Gender section. Civil ) level categorized 26% of QDS users 2

1 Service Management and African

0 as non-poor. The low proportion is 2 Institute for Capacity Development d likely due to bad weather during e t (AlCAD) for funding the MA a the reference year. These results d ( emphasis that climate (especially programme and fieldwork that is r e h rainfall) in semi-arid conditions the basis of this paper. Support s i l from all individuals during b influence Significantly the u

P fieldwork is highly appreCiated.

perfonnance of on-fann seed e h t production. y b

References d e t Recommendations n a r Production and use of gDS should Collier.P .. Radwan.S .. Wangwe. S. g

e and Wagner. A. 1990. Labour

c be encouraged and supported by n e the agricultural departments of the and Poverty in Rural Tanzania c i l

(fjamaa and Rural r district councils both technically e d and finanCially to improve Development in The United n u agricultural production and Republic of Tanzania. y a Clarendon Press. Oxford.

w productivity to reduce poverty in e t 143pp.

a tenns of food security and income. G

t Delibera te efforts by Local e

n Mbwele. A. A.. Lumbadia, M.Z. i Government AuthOrity to b a coordinate different institutional andSichilima, N.P. 2000. Seed S

y production and supply system

b efforts that support gDS is

d in Tanzania: In: Proceedings of e required. The activities that need to c u be coordinated with regard to QDS the Stakeholders Review and d o r include: Production that reflects Planning Workshop. 7-8 p e R Seed Production and Proverty Reduction 171

December 1999, Dar es %20SUMMARY.doc.j site Salaam. Tanzania. pp20-27. visited on 17/07/2004.

Muliokela, S.w. 1997. Seed URT. 1994. National Seed Policy supplyconstraints in Southern and Implementation and Eastern Africa: Alterative Guidelines. Ministry of strategies for smallholder seed Agriculture. 40pp. supply. In: Proceedings oj an International Conference on URT. 1999. Poverty and Welfare Regional Seed Systems in Monitoring Indicators.: Vice­ AJrican and West Asia Edited president's Office Dar es by Rohrbach, D.D.et al. 10-14 Salaam, Printed by the March 1997, , Government Printer, Dar-es­ , pp. 11-19. Salaam. 76pp.

TARPII-SUA. 2002. Impact oj URT. 2001a. Rules, Regulations agricultural Research: Done in and Procedures for Quality the pas t Twenty Years Jor Declared Seed Prod uction in Selected Commodities in the Tanzania. Ministry of Eastem and Southem Agriculture and Food Security. Highlands Zones, Tanzania. Dar-Es-salaam University

) TARPII-SUA Project, Press, Dar-es-salaam. 71 pp. 2

1 Tanzania. 126 pp. 0 2 URT. 2001 b. Agricultural Sector d e t

a Temu, A.A .and Mtenga, K.J. 2000. Development Strategy d (

Situation and outlook for the [http://www.kilimo.go.tz/publ r e

h development of seed icationsj.site visited on s i l

b production and marketing 15/03/2004. u

P systems in Tanzania; e h URT 2002. Poverty and

t Implication and distribution y

b strategies: Bean cowpea humanDevelopment Report. d e collaborative research support Mkuki na Nyota Publishers, t n

a programme . In: Dar-es-salaam, r g Proceedings Bean Seed Tanzania.113pp. e c n Workshop. 12-14 January e c i l 2000, Tanzania, pp 1- URT. 2003. Poverty Reduction r e 6. Strategy: The Second Progress d n

u Report 2001/02. Government y a UnHed Nations. 2002. Economic Printers. Dar-Es-Salaam, w e t and Social council Economic Tanzania 92pp. a

G Commission for Africa. Joint t e ECA/OAU / ADB secretariat in n i b collaboration with UNFPA. a S [http:/www. uneca.org/ eca y b

d resource/major ECA e c websites/icpd/fourth/FC4- u d o 03%20doc%20demo%20Trans r p e R