The adoption and impact of an improved groundnut variety in mixed crop-livestock systems in Southern India

Transcription

1 The adoption and impact of an improved groundnut variety in mixed crop-livestock systems in Southern India Nils Teufel, Nancy Johnson and Dhiraj Singh Asian Society of Agricultural Ecomics Oct 13-15, Hai

2 Outline Background on study site and varietal development program Sample selection Characteristics of households Farming systems Groundnut varietal use Discussion Impact of variety on net revenue and productivity Discussion

3 Background-Anantapur Pop 3.6 million; 75% rural Named among 250 most backward districts in 2006 Groundnuts planted on 80% of cropped area (800,000 ha) Yields are low:.27t/ha) Average rainfall: 550mm/year Livestock are important for income and coping with risk, and groundnut haulms are a major source of feed ANDHRA PRADESH ANANTAPUR

4 Despite the importance of groundnut, farmers have been reluctant to replace their traditional variety (TMV2 ) with improved cultivars was developed by ICRISAT, ILRI and partners for drought-tolerance, yield (pods & haulms) and quality (haulms) Officially released in 2006, based on data from research stations and farmers fields Advantages compared to TMV2 15% increase in pods and haulms Cows fed on gave 0.5 lt more milk per day

5 Promoted by Accion Fraterna, a local NGO involved in the varietal development, via farmer-to-farmer sales Estimates of dissemination to date: 285 ha in ,000 ha to 12,000 ha in 2009 (1.5% total area) 25,000 ha in 2011 (3.1% total area)

6 Objectives of the study Estimate the adoption and impact of on crop and livestock productivity Assess gender and intra-household implications Understand adoption pathways and dynamics Assess influence of crop-livestock interactions on breeding and dissemination

7 Location of study and sampling frame Guntakal Gooty Vidapanakal Yadiki D.Hirehal Vajrakarur Peddavadugur Bommanahal Pamidi Tadpatri Peddapappur Uravakonda Kanekal GarladinneSinganamala Rayadurg Gummagatta Kilometers Beluguppa Bukkaraya Samudram Kudair Putlur Anantapur Narpala Yellanur Atmakur Vill survey hamlets VS random Icgv (list) VS random n-icgv (list) VS added for Icgv Anantapur mandals other AF - dry AF - wet NonAF - dry NonAF - wet BrahmasamudramKalyandurg Tadimarri Raptadu Settur Kambadur Kanaganapalle Bathalapalle Mudigubba Dharmavaram Ramagiri Kundurpi Chennekothapalle Bukkapatnam Talupula Kothacheruvu Nallamada Amarapuram Kadiri Roddam PenukondaPuttaparthi Gandlapenta Nambulipulikunta Obuladevaracheruvu Gudibanda Somandepalle Gorantla Nallacheruvu Madakasira Amadagur Parigi Tanakal Rolla Chilamathur Agali Lepakshi Hindupur

8 Adoption of based on Village Census hamlet category random selection purposive naf AF mandal AF mandal selection mandal ICGVhamlehamlehamlet n ICGV- ICGV- nicgvhamlet hamlet [.] share of groundnut area [%] area [ha] share of groundnut growing hh [%] growing hh [.] Source: Village census

9 Mean values of selected household characteristics in sample hamlets Adoption status of ICGV Adopters Non-adopters, adopting hamlets Non-adopters, n-adopting hamlets Households [.] Age Hh head 1 [y] Hh size [.] Land owned [ha] Land cultivated [ha] Groundnut area [ha] area [ha] Dairy cattle, local [.] Dairy cattle, crossbred [.] Buffalo [.] Donkey [.] Goat [.] Sheep [.] Draft animals [.] Source: Village census

10 Results of sample selection using propensity score matching (PSM) Adopters n-adop., adop. Hamlet n-adop, n-adop. hamlet LandCultAc unmatched ATT LandOwnAc unmatched ATT CtlLcl unmatched ATT Draft unmatched ATT

11 Location of selected hamlets by household survey category Hh survey hamlets HS adopting hamlet HS n-adopting hamlet Non HS hamlet Anantapur mandals AF - dry Antharaganga AF - wet Kammuru Nakkalapalli Brahmanapalli NonAF - dry Narinja Gundlapalli Talupur NonAF - wet VenkatadripalliDuddekunta P. KothapalliKurlapalli Upparapalli Yalagalavankathanda Bukkacherla Vepacherla Eguvathanda Ramanepalli KadadarkuntaThumucherlaYerragunta Eradikera Gollapalli Chinnampalli Duradakunta Upparalapalli KamireddipalliGollapalliThummala Ayyagarlapalli Regatipalli Thimmapuram Mulakaledu Tenegal Chennampalli Tanakontivaripalli Yerraguntla Appilepalli Malayanur Kilometers

12 Mean value of selected demographic characteristics and assets Adopters (n=80) Non-Adopter AH (n=91) Non-Adopter n- AH (n=102) Mean Mean Mean Age of hh head [y]** Edu of hh head [y schooling]*** Hh size [.] Labor available (15-59 years) [.] Cultivated land owned [ha] Cultivated irrigated land [%]*** Cattle local [.] Cattle xbreed [.]*** Buffalo local [.] Buffalo improved [.]*** Total Adult animals [TLU] Non-agri asset index Agri asset index Source: Household survey

13 Mean value of selected demographic characteristics and assets Adopters (n=80) Non-Adopter AH (n=91) Non-Adopter n- AH (n=102) Mean Mean Mean Age of hh head [y]** Edu of hh head [y schooling]*** Hh size [.] Labor available (15-59 years) [.] Cultivated land owned [ha] Cultivated irrigated land [%]*** Cattle local [.] Cattle xbreed [.]*** Buffalo local [.] Buffalo improved [.]*** Total Adult animals [TLU] Non-agri asset index Agri asset index Source: Household survey

14 Mean values of key performance and welfare measures of households Non-Adopter AH Non-Adopter Adopters (n=80) (n=91) n-ah (n=102) Milk production [l/d]*** Milk marketed [%]*** Annual income [INR] 173, , ,028 Crop income [INR] 125,886 63,987 62,626 Livestock income [INR] 25, ,307 Annual Food Expenditure [INR]** 36,994 32,342 32,716 Annual Non-Food Expenditure [INR] 65,514 63,008 58,955 Non-agri asset index Agri asset index 2011*** Change in agri-assets index*** Change in n agri-asset index* Source: Household survey

15 Mean values of key performance and welfare measures of households Non-Adopter AH Non-Adopter Adopters (n=80) (n=91) n-ah (n=102) Milk production [l/d]*** Milk marketed [%]*** Annual income [INR] 173, , ,028 Crop income [INR] 125,886 63,987 62,626 Livestock income [INR] 25,045 19,751 13,307 Annual Food Expenditure [INR]** 36,994 32,342 32,716 Annual Non-Food Expenditure [INR] 65,514 63,008 58,955 Non-agri asset index Agri asset index 2011*** Change in agri-assets index*** Change in n agri-asset index* Source: Household survey

16 Distribution of area and households under different crops by households status in that season planted Rabi ( ) Kharif (2010) planted (nadopting (adopting (adopting area hh area hh area hh area hh 97ha n=71 92ha n=70 319ha n=56 598ha n=115 (nadopting area hh 492ha n=102 area 29ha hh n=15 [%] [%] [%] [%] [%] [%] [%] [%] [%] [%] [%] [%] Groundnut Paddy Other cereals Pulses Fruits Vegetables Other

17 Distribution of area and households under different crops by households status in that season planted Rabi ( ) Kharif (2010) planted (nadopting (adopting (adopting area hh area hh area hh area hh 97ha n=71 92ha n=70 319ha n=56 598ha n=115 (nadopting area hh 492ha n=102 area 29ha hh n=15 [%] [%] [%] [%] [%] [%] [%] [%] [%] [%] [%] [%] Groundnut Paddy Other cereals Pulses Fruits Vegetables Other No major differences in cropping pattern in Karif

18 Distribution of area and households under different crops by households status in that season planted Rabi ( ) Kharif (2010) planted (nadopting (adopting (adopting area hh area hh area hh area hh 97ha n=71 92ha n=70 319ha n=56 598ha n=115 (nadopting area hh 492ha n=102 area 29ha hh n=15 [%] [%] [%] [%] [%] [%] [%] [%] [%] [%] [%] [%] Groundnut Paddy Other cereals Pulses Fruits Vegetables Other Adopters plant more groundnut in rabi

19 Area and household distribution of various groundnut varieties by households status in that season planted ICGV Rabi ( ) Kharif (2010) planted (nadopting (adopting (adopting (nadopting area hh area hh area hh area hh area hh area hh 18ha n=15 34ha n=30 20ha n=20 268ha n=56 481ha n= ha n=102 [%] [%] [%] [%] [%] [%] [%] [%] [%] [%] [%] [%] ICGV TMV (local) K JL Kharif adopters plant much less TMV2 than n-adotpers and a little less K6 AF villages more K6

20 Area and household distribution of various groundnut varieties by households status in that season planted ICGV Rabi ( ) Kharif (2010) planted (nadopting (adopting (adopting (nadopting area hh area hh area hh area hh area hh area hh 18ha n=15 34ha n=30 20ha n=20 268ha n=56 481ha n= ha n=102 [%] [%] [%] [%] [%] [%] [%] [%] [%] [%] [%] [%] ICGV TMV (Local) K JL Rabi More improved varieties; adopters plant mostly ICGV

21 Grain and haulm yield [t/ha] of different crops by households status in that season planted Rabi ( ) Kharif (2010) planted (nadopting (adopting (adopting (nadopting grain haulm grain haulm Grain haulm grain haulm grain haulm grain haulm ICGV TMV (local) K JL Kharif better than TMV2 and usually better than K6 Rabi t as clear

22 Source of seed by variety (%) ICGV K6 (n=69) TMV2 (n=206) JL24 (n=8) (n=80) Farm saved Exchange Purchase from other farmers Purchase from market Provided by promoting agency/ngo Provided by govt. organisations

23 Source of seed by variety (%) ICGV K6 (n=69) TMV2 (n=206) JL24 (n=8) (n=80) Farm saved Exchange Purchase from other farmers Purchase from market Provided by promoting agency/ngo Provided by govt. organisations

24 Distribution of households [%] reporting links to AF 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Adopter Non-adopter (n-adopting village) response link yes

25 Distribution of households [%] reporting a role of institutional support in their decision on groundnut variety 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Adopter Non-adopter (adopting village) Non-adopter (n-adopting village) All (n=273) response yes

26 Pods Use of pods and haulms Most are sold to traders to be processed for oil Some are sold for seed, kept for seed and used for home consumption Price premium for ICGV9114 Haulms Most neither purchased r sold Of those who did, more purchased than sold No price premium for haulms Haulm purchases and sales adopters n-adopters n-adopters (adopting (n-adopting purchased sold purchased sold purchased sold quantity [t] price [rs/t] 1,990 1,429 1,728 1,596 1,569 1,688 hh [.]

27 Value of output, yield and cost of cultivation of, TMV2 and K6 by season Rabi ( ) Kharif (2010) ICGV ICGV TMV2 K TMV2 K6 Hh growing [.] Outputs Value of production[rs/ha] 64,398 26,835 52,480 21,810 7,440 17,014 pod yield [t/ha] haulm yield[t/ha] Input costs [Rs/ha] seeding/planting seed 7,310 5,221 7,741 5,032 3,850 5,268 irrigation fertiliser 1,079 2,066 2,874 1, ,263 manure 1,354 1,529 1,813 2,057 1,282 1,917 weeding 1,746 1,977 1,632 1,522 1,229 1,757 pesticides 1,098 1,106 1, harvesting/threshing 3,081 2,002 2,577 1, ,485 other Total [Rs/ha] 16,447 14,617 18,731 12,704 8,938 13,001 Net revenue[rs/ha] 47,951 12,217 33,749 9,107-1,499 4,013

28 Results of production function estimation (dependent variable: value of production-pods + haulms, n=421) Independent variables Production costs Sign and significance Irrigation, fertilizer +++ Planting, seed, manure, weeding, other Demographics (age, education) Land cultivated -- % irrigated land cultivated +++ Rabi dummy +++ TMV dummy --- K6 dummy JL24 dummy -- Non-adopter in adopting hamlet Non-adopter in n-adopting hamlet

29 Results of production function estimation (dependent variable: milk yield (litres/cow/day; n=175) Independent variables Sign and significance Purchased fodder ++ Purchased concentrate % of milk marketed +++ % Buffalo in herd % Improved animals in +++ herd Demographics (age, education) Land Cultivated % irrigated land cultivated Non-adopter in adopting hamlet Non-adopter in nadopting hamlet

30 Conclusions Adoption of generates significant ecomic benefits, including to small farmers, yet despite recent growth, few households are adopting it. Some possible reasons include: Seed systems. The fact that it is t in the public distribution system makes it more costly to obtain, however a significant number of farmers get other improved varieties from farmer-to-farmer dissemination, and seed cost is t higher for compared to K6. Lack of links to public credit and insurance system cold be more important (see below) Early stages of dairy intensification. Evidence suggests that feed quality only becomes important in later stages of intensification process. The haulm characteristics are t yet widely appreciated; farmers plant it for pod yield and market doesn t reward quality traits. Role in risk management. is drought tolerant, but for vulnerable farmers seeking to minimize risk this is a difficult message to communicate,and in any case they might be better off relying on government crop insurance systems. Some implications Strengthening farmer to farmer system could encourage dissemination of and others improved varieties Different characteristics of the variety might appeal to different audiences=> targeted dissemination eg to dairy farmers