Journal of Developments in Sustainable / : ,,/ῌ ,-3 ( ,*+* )

Adoption of Modern Agricultural Technology by the Koya People, an Ethnic Minority in

Hiroyuki Tsunashima*

Research Center for Urban Safety and Security, Kobe University, Rokkodai, Nada, Kobe,0/1ῌ 2/*+ , Japan

Ethnic minorities in India mostly live by agriculture in hilly forested areas. Despite the various e# orts of the government, the response of the ethnic minorities in modernizing agriculture has been rather sluggish. The greatest concern of the past research on ethnic minorities in India was how they could adopt mainstream agricultural technologies as a complete package. However, a field survey in a village of the Koya people revealed that they have a di# erent idea of agricultural development. In this regard, this elucidates how they have interfaced modern technologies with their traditional cropping systems. The results show that the traditional cropping systems consist of invisible but knowledge intensive tasks such as decisions on fertilization and choice of /variety. During the survey period, an increasing number of villagers started cash cropping with large amounts of chemical and pesticides. Simultaneously, land lease contracts related to cash cropping have also been common. This paper also discusses the significance of this new trend in . The high-input system for cash cropping is costly and exposed to the mainstream market economy. Thus, cash cropping is accompanied by the risk of sudden decline in the selling price. The data shows that the land lease system has a function of redistributing cash income within a village community, and that the Koya people view land lease in a di# erent light than reported in past research. In conclusion, this paper suggests that the Koya people can be a good example of a realistic alternative to mainstream agricultural development.

Key words: agroecological perception, cash income, crop choice, indigenous knowledge and technology, land use

ῌῌῌῌῌῌῌῌῌῌῌῌῌῌῌῌῌῌῌῌῌῌῌ

Introduction improved seeds and chemical fertilizers could not be e# ectively utilized, because the topography was Ethnic minorities in India called “scheduled unfavorable for , and moreover, small- tribes” live mostly by agriculture in the hilly fore- scale land holding discouraged e# orts to install sted areas with limited access to natural resources, irrigation facilities. Several succeeding studies con- information and markets (cf. Mehta and Shah, ducted on di# erent ethnic minority groups reported ,**-; Shah and Guru, ,**- ). Consequently, they that governmental agencies still almost always fail have lagged behind mainstream agricultural devel- to disseminate mainstream agricultural technol- opment (cf. Thakur and Thakur,+33. ; Gulati and ogies among the ethnic minorities (cf. Hansda, Kelley,+333 ; Guilmoto, ,**, ). +33. ; Iyengar, +33. ; Kornel, ,**0 ). The greatest Past research on agricultural development for eth- concern of the past research was how ethnic mi- nic minorities has discussed this lag in detail. norities could adopt mainstream agricultural tech- Investigating the socioeconomic condition of the nologies as a complete package. Today, however, it Koya people, Ramaiah (+32+ ) made one of the has become obvious that the simple transfer of the earliest systematic studies on agricultural develop- package is not realistic in an area where ethnic ment for ethnic minorities in India. He argued that minorities live.

Received: April, , ,*+* , Accepted: December +* , ,*+* * Corresponding author: Research Center for Urban Safety and Security, Kobe University, Rokkodai, Nada, Kobe, 0/1ῌ2/*+, Japan. Tel:ῌῌ2+ῌῌ 12 2*- ῌ 0,/0 , Fax: 2+ ῌῌ 12 2*- ῌ 0*,2 , E-mail: [email protected] 226 J. Dev. Sus. Agr./, ( )

During the period when I was making an investi- gation into indigenous knowledge and technology of the Koya people, I observed farmers to be adopt- ing some modern technologies according to their own choices. Also I found a new trend in land use that followed the adoption of these technologies. I was convinced that the case of the Koya people could be a good example of a realistic alternative. Accordingly, on the basis of my own investigation in a village of the Koya people, the present paper elucidates how they have interfaced modern tech- nologies with their own cropping systems. In this paper, existing cropping systems are firstly outlined with the extent to which the modern tech- nology has been adopted. Secondly, the relation- ship between the technology adoption and agro- ecological perception of the Koya people is de- Fig.+ . Location of the study area. scribed. Thirdly, the new trend in land use is re- vealed and its significance in future agricultural development is discussed. In conclusion, this paper provides a di# erent perspective from past research on economic activities of ethnic minorities. Methods

Study area The Koya people are viewed as a typical sched- uled tribe (ST) as they live mainly by agriculture and forest product collection. They reside in hilly forested areas along the Godawari River in the northern inland part of State (Fig. +). This area, where three Statesῌ Andhra Pradesh, Chattisgarh and Orissaῌ border each other, has a significantly denser population of various STs than any other part of South India. District has the highest ratio of STs to the total population Fig., . Topography of Chalampalem. of Andhra Pradesh (,/ῌ ). The Koya people have The contour lines indicate altitudes of.* m, 0* +** the second largest population (-/, , *** ) among -- m and m. The gray zones with broken lines indicate reserved forest. The hemmed line indi- STs in Andhra Pradesh, and the greatest population +32+ cates a road. in , according to the Cen- Source: Government of India (+321 ). sus (Subramanyam,,**- ). The fieldwork was conducted in a village of the Koya people in Mandal. The study chili-trading center in Chattisgarh State. village, named Chalampalem, is an administrative Data collection village consisting of three hamlets called Erragattu, The fieldwork was carried out from December Chalampalem and Bandarigudem (Fig., ). All ,**, to August ,**- and from September ,**. to households in the neighboring administrative vil- April,**/ . At the beginning, I explored the entire lage to the south, Nandigama, are made up of non- study village with a mobile GPS to draw a farmland tribal people. A paved road goes from Chalamp- map. It showed uneven spatial distribution of alem through Nandigama and leads to Jagdalpur, a cropping patterns. Then I surmised that soil condi- Tsunashima: Modern Agricultural Technology and the Koya People 227 tions should a# ect crop choice of farmers and they to every crop that interviewees grew in,**, , they might have their own soil taxonomy. were asked what they considered a good yield to be. The investigation of villagers’ soil classification In April and May,**/ , the second all-household was undertaken by transect walking through the survey was carried out in the same manner as the site guided by a villager. He showed me several soil first all-household survey. A farmland map was types and their characteristics. Later, based on the also drawn again. results of the sample household survey below, I Throughout the fieldwork period, observation of specified the two main soil types in the study vil- farmwork was also conducted whenever necessary. lage. In June,**- , I observed soil profiles by digg- Implements and materials used were recorded for ing three observation pits for each of these main each task. Farmers were interviewed about the soil types along the transect line. necessity and significance of each task in an infor- In order to collect data on the present situation mal manner. Informal interviews with particular of each cropping system, it was necessary to con- key informants were also held whenever necessary. duct a household survey with questionnaires. Be- Information about the installation and purchase fore this, with the aim of briefing myself on the costs of irrigation implements and wages of agricul- current change in agriculture, I conducted group tural laborers was collected. interviews in each hamlet on di#$ erent days in For all interviews, I used Telugu, the o cial March,**- . The participants were +* to ,* resi- language in Andhra Pradesh, with additional assis- dents of each hamlet and members of a local NGO tance from young villagers. engaged in rural development activities based in Data analysis Bhadrachalam. In the group interview, we dis- Data collected through the household surveys cussed the-* -year pattern of changes in cultivation was sorted by cropping system so that di# erent methods, particularly with regards to the introduc- cropping systems could be compared with each tion of the modern technology. In this regard, the other. The data on farmers’ behavior was used for participants mentioned a recent trend toward land identifying the extent to which the modern technol- lease with an economic motivation. ogy had been adopted. The first all-household survey was conducted in In order to provide an overview of farmers’ agro- April and May,**- . I visited each household to ecological perception, the principle of the villagers’ interview members who were playing a central role soil classification was firstly inferred from the ob- in the farmwork. I collected information on family servation of the soil profiles. Secondly, the data structure, holding, agricultural relating to rationale behind the farmers’ choice of implements held, held, land tenancy, cropp- crop and technology was compiled. As the house- ing area and yields for every crop grown in the,**+ hold surveys were not su $ cient, the narrative data and,**, cropping year. Further, if all or a portion collected through the informal interviews was used of the harvest was sold, I also collected information to make up for the deficit. This data explained the on the amount of sales and the selling prices. crop choice according to subtle soil conditions and Subsequently, the sample households were cho- the rejection of some technologies. sen for a further intensive survey to investigate the The new trend toward land lease was analyzed cultivation methods of+* major grown in the with the data collected through the all-household case study village. The sample household survey surveys. Firstly, an increase in area under land was conducted in June and July,**- . For every lease and number of cases was demonstrated. Sec- crop, information on the following items was col- ondly, according to the suggestion given in the lected: varieties adopted, suitable , timing group interview, a comparison among revenues and of tasks such as tillage, weeding, inter-tillage, fertil- land rents of several crops was made. Then, the izer application, spraying pesticide, required fre- motivation and the meaning of land lease were quency of these tasks, costs and amount of applying analyzed. manure, chemical fertilizers and pesticides. The In the text, words in italics indicate local terms. reasons for choice of each crop and adoption or The currency unit is rupee (abbreviated to Rs.). Its rejection of each technology were also inquired. As exchange rate versus the U. S. dollar was Rs. .,ῌ ./ 228 J. Dev. Sus. Agr./, ( ) during the study period. season. The villagers describe this phenomenon as Results and Discussion “the coming and going of the Godawari”. Crops Outline of Agriculture in the Study Village The staple crop is currently wetland rice, which Agrarian structure is raised only in the rainy season. The main rainy- As of April,**- , there were +22 households or season upland crops are upland rice, red gram, 2/0people in Chalampalem, among which +2- cotton, little millet and pearl millet. Black gram households or2-2 people were the Koya. Almost and green gram are planted in the post-rainy sea- all the Koya households made their livelihood by son. Sesame is grown in the both seasons. The farming or agricultural wage labor. In all,+./ cropping season for sorghum depends on variety. households were engaged in farming, while the As for chili, nurseries are prepared in the rainy others were exclusively made up of wage laborers season and seedlings transplanted at the beginning except for a few households managing small gener- of the post-rainy season. al stores in their residences. Of the total number of Today rice has become the main staple for the Koya households,,* percent had no cultivated villagers. Traditionally, they have been also eating land, and the operating acreage of another1* per- a range of millets, among which sorghum has been cent was less than, ha. Five non-tribal households especially important. According to elder villagers, also managed small general stores and had no farm- various species of coarse millets used to be cultivat- land. Some non-tribal farmers from Nandigama ed until+32* s. are usually used for curry rented considerable areas of land for chili cultiva- and served with the staple . Cooking oil is tion from Chalampalem Koya villagers and hired squeezed from sesame. Residue of rice, millets and them for agricultural wage labor. legumes are important as fodder. Agroecological conditions Formerly, mixed cropping of sorghum and green The western boundary of Chalampalem is a gram was dominant in Chalampalem. However, Godawari tributary, which runs southward in the the mode of agriculture is undergoing a remarkable rainy season but dries up in the dry season (Fig., ). shift: various cropping patterns with millets have The north and east fringes of Chalampalem are become rare, while chili has become the most ec- hilly deciduous teak-bearing forest. Settlements onomically important crop, followed by cotton and most agricultural fields are located in the flat (Table+, and ). Chili and cotton are produced for zone. Shifting cultivation fields are sometimes sale only, while rice and sorghum are mostly for opened in the hilly forest zone. Palmyra palm trees home consumption. line the boundary of individual agricultural land Adoption of modern technologies holdings and large tamarind trees planted near The extent to which the modern technology has houses are prevalent. The rest of the flat zone is been adopted varies according to crop (Table- ). grassland often utilized as lands. This flat Chili cropping is a complete adoption of the mod- zone is dotted with large and small rocky hills ern technology package, which includes lift irriga- (inselberg). tion, tillage with , purchased seeds of mod- The nearest climate observatory is located in ern varieties, chemical fertilizers and pesticides. Bhadrachalam (,* km west of Chalampalem, N Modern varieties of cotton are cultivated in rain- +1ῌῌ -3ῌῌ, E 2* /, ). Most of the annual rainfall oc- fed conditions; no traditional variety is raised. curs from June to October. Annual rainfall dras- These cash cropping systems are accompanied by tically changes year by year. The distribution pat- the most intensive application of chemical fertiliz- tern of annual rainfall is also variable. ers and pesticides. The level of the Godawari annually fluct- Various modern varieties of wetland rice have uates according to rainfall in its vast watershed, been introduced. Paddy fields have been fertilized and normally reaches its highest level in the middle with dung as a basal application in a local of August. At that time, a considerable part of the manner. Today, villagers have also come to top- land along the tributary is submerged for some dress their paddy fields with chemical fertilizers. days, which announces the coming of the post-rainy Pesticide spray has also become common. All wet- Tsunashima: Modern Agricultural Technology and the Koya People 229

Table+ . Change in cropping area and number of cultivator households by cropping pattern

Year,**, Year ,**- Year ,**. Cropping pattern No. of No. of No. of Area (ha)cultivator Area (ha) cultivator Area (ha) cultivator Sorghum῍ Green gram .240 1. +.4+ -* 34- ,+ Wetland rice .24, 23 0*43 +++ 0/4. ++1 Chili ,.4* ,1 -/4/ // .140 10 Red gram+ῌ +*42 ,. 14/ +1 .4+ 3 Upland rice῍ Red gram 04/ 1 /4, +* -40 1 Cotton+ῌ /4* +- 34* ,* +34- .2 Sesame .4* 1 ,4, 0 14, +/ Upland rice ,42 - *4* * *4. + Red gram῍ Sesame ,42 - *4* * *4* * Sorghum῍ Pearl millet ,4* , *4* * *4* * Sorghum +43 . +42 / *4* * Black gram +42 / *4* * *4, + Green gram *4* * *4, + *42 - Sorghum῍ Sesame +40 . *4* * *4- + Sorghum῍ Black gram +4, - *42 , *4* * Sorghum῍ Little millet +4, , .41 0 *4* * Eucalyptus +4, + ,4, , -4. - Red gram῍ Green gram *40 , *4* * *4* * Green gram῍ Sesame *4* * *4* * *40 - Total +0.4, +,2 +..4+ +,2 +0,4, +-, Source: all-household survey conducted in Aprῌῌ May,**- and Apr May ,**/ . +ῌ Including mixed-cropping with maize or little millet. land rice varieties were grown with tank-irrigation sesame, legumes and millets (Table- ). This mod- which entirely depends on rainfall. est application of modern inputs is what past re- The other crops are all traditional varieties and search on ethnic minorities in India regarded as a grown under rain-fed conditions. Upland rice re- problem. ceives as much chemical as wetland rice. Villagers unanimously said that fertilization with Sesame is ordinarily fertilized with cow dung every both chemicals and manure will allow weeds to three years, because repeated cultivation without flourish and consequently invite pest insects. Some fertilization gradually decreases yields. Sorghum villagers further explained that the fertilization of and other millets are cultivated in the traditional millets itself also attracts pest insects. It naturally manner without any fertilization or pest control follows that, if they are to use some amount of measures. fertilizers, they have to repeat weeding; if they are to apply a larger amount of fertilizers but have no Relationship between Technology Adoption and time to repeat weeding, they must also be ready to Farmers’ Agroecological Perception purchase pesticides. Farmers’ perception of manure and chemical fertil- A comparison of cultivation methods among sev- izers eral cropping patterns shows that those cropped Villagers began to use chemical fertilizers and with a lot of fertilizer require more frequent weed- pesticides as early as the+31* s or +32* s. However, ing and pesticide spray (Table- ). Traditional cropp- they still apply only small amounts of chemical fer- ing systems with sorghum require almost no pesti- tilizers and pesticides to traditional crops such as cide spray. Villagers intentionally avoid fertiliza- 230 J. Dev. Sus. Agr./, ( )

Table, . Land productivity of main cropping patterns in the study village

,-.ῌῌῌ +ῌ Yield Selling price Gross return Cropping pattern Year (kg/ha) (Rs/kg) (Rs/ha) ,**+ -+1῍῍ +0+n.a. +/4,+ ,..3 ῎ ,**, +,0῍῍ ./ +4/* +/4/1 23* Sorghum῍ Green gram ,**- ,2,῍ +.2 n.a. n.a. ,**. .0῍ +2 n.a. n.a. ,**+ ,*3. .413 +**-* ,**, 23 n.a. n.a. Wetland rice ,**- ,2-/ -4/3 +*+,+ ,**. ++/, /4** /10* ,**+ -*+3 -.4/0 +*.--1 ,**, ,2,1 -/4-3 +***.2 Chili ,**- -*0- ,2403 21220 ,**. ,.1* ,+433 /.-*0 ,**+ +0+ +,41+ ,*.0 ,**, 30 +-4** +,.2 Red gram ,**- ,.3 +-4** -,.+ ,**. ,.+ n.a. n.a. ,**+ ,3-῍῍ +.+ -41/ +,41+ ,23+ ,**, +,.῍῍ 1-n.a. +-4** 3.3 ῎ Upland rice῍ Red gram ,**- /3*῍῍ +-1n.a. +-4** +111 ῎ ,**. +/+῍ 3/ n.a. n.a. ,**+ 013 +/4.- +*.11 ,**, --, +143. /3/0 Cotton ,**- 31, +24/* +131- ,**. 2,. +/4.2 +,1/-

,**+ n.a. n.a. n.a. ,**, ,.1 +-401 --10 Sesame ,**- +-0 n.a. n.a. ,**. +-0 +34/* ,020

Source: all-household survey conducted in Apr - May,**- and Apr - May ,**/ . +ῌ Cropping patterns adopted by+* or more households in either year during ,**,ῌ ,**. are listed. ,ῌ In cases of inter-cropping or mixed-cropping of two crops, yields of the both crops are presented -ῌ Prices received by producer for chili and cotton, farmyard prices for the other crops. Rs: rupees. .ῌ The value of mean gross income is calculated by multiplying the yield by the selling price. In cases where the selling prices of the crop(s) are unknown because there was no sale, the value is treated as “not available”. However, in cases where the selling price of one crop in a multiple cropping pattern is unknown, the value of the other crop alone is presented with “῎ ”. tion for the purpose of preventing the rampant not only frequent hand-weeding but also regular spread of weeds and consequently controlling pest planting which enables the eradication of weeds by insects without spraying pesticides. On the other inter-tillage with ploughs. Still, pest control costs a hand, cash crops such as chili and cotton require considerable amount of money in the extreme. Tsunashima: Modern Agricultural Technology and the Koya People 231

Table- . Comparison in amount of fertilizer inputs and the frequency of weeding and pesticide spraying among various crops and wetland rice varieties in an average year

Chemical fertilizers Pesticides Times /ῌ Planting Irrigation ManureCosts Times Costs Crop/variety Nsystem01ῌῌ conditions (t/ha)NPK of of (Rs/weeding (Rs/ (kg/ha) appli- 2ῌ ha)cation ha) Chili+, R, TP Lift * .02 ,-. ,2 +*,02 -4.3ῌ +*4* ++.01 Cotton. R, DS Rain-fed +401 -+0 +,3 * 0,33 -4-3ῌ -40 ,/31 Wetland rice BPT-/,*. (late)+ῌ . IR, TP Tank .4.* ,,0 ++1 +0 /*/- ,4* ,4* 1,2 MTU-1*,3 (medium) ,, IR, TP Tank ,4-* +22 3, + .+32 ,4+ +41 0*2 IR-0. (early) ++ IR, TP Tank +4.* +0, 1+ * -/-1 +42 +4/ /0- Upland rice,ῌ . IR, DS Rain-fed ,4,, +0- .3 * -*+- ,4- *42 ,1. Sesame0 IR, DS Rain-fed +43+ ,- . * -/. *42 *4/ +2, Red gram+- IR, DS Rain-fed *4-1 0 , * ++0 *41 *4, /0 Sor῍ G-ῌ -, IR, DS Rain-fed *4*/ . - * +++ *40 *4* +1 Sor῍ M.ῌ 1 IR, DS Rain-fed * * * * * +4, * * Source: sampled-household survey conducted in May - June,**- . +ῌ Maturation characteristics of wetland rice varieties. ,ῌ Including inter-cropping with red gram. -ῌ Sorghum and green gram or black gram. .ῌ Sorghum and little millets or pearl millets. Cultivated in slash-and-burn method. /ῌ Number of sample households. 0ῌ R: regular planting, IR: irregular planting, TP: transplantation, DS: direct sowing. 1ῌ Air-dry weight. In many cases manure is self-supported. 2ῌ The values were estimated based on the information from a few interviewees about types of pesticides and amounts of spray. 3ῌ Times of inter-tillage, which is followed by hand weeding.

Traditional method of manure application this a household on average obtains+, . -t of manure The Koya people have been applying manure annually. These methods are mainly adopted for with their cattle in the following manners. Some cultivation of sesame and wetland and upland rice farmers corral or tether their own cattle at a place (Table- ). in their plot after harvesting. Remaining at that Traditional method of optimizing output other than place overnight, the cattle repeatedly tread on their fertilization dung and automatically fertilize the plot with the As mentioned earlier, millets are raised without smashed dung. The farmers move the corrals or any fertilization. Here, a question arises as to what piles once every few days until the entire plot has has been the traditional method of increasing the been uniformly fertilized. This method is adopted yield other than fertilization. According to villag- only in the cool season, namely from December to ers, preparatory plowing and timely sowing is in- February. In the agricultural-o# season, it is too stead a method of increasing yield for the mixed- hot for cattle owners to water the cattle in their cropping of “yellow” sorghum (a sorghum variety) fields. Other farmers tether their own cattle near and green gram in the post-rainy season. It is said their homestead every night to produce smashed that the more frequently the land is plowed, the dung and bring it to their fields at the beginning of better the yield. Hard-working farmers plow the the cropping season. On average, a cow or bullock land twice or three times even before “the Godawari produces*,0 . -t of dung in air-dry weight, and from comes” to facilitate soakage of subsurface soil with 232 J. Dev. Sus. Agr./, ( ) rainwater. After the soil surface becomes moder- ately dry, they again plow around four times after “the Godawari has gone”. The combination of these two crops itself is also a form of technology. This combination is a matter of course for most villagers. Only a few villagers explained that the merits are shading of green gram by the tall sorghum and covering of soil surface by the creeping green gram, both of which retard the soil surface drying up in the post-rainy season. Besides this, various mixed cropping were ob- served. The most common was combinations of fast ripening crops such as upland rice and little millet (growing period:- months) and slow ones such as “hill” sorghum (another sorghum variety), red gram and cotton (growing period:0 months). Some villagers explained the merit of those cropp- ing patterns, saying that such crop combinations make it possible to harvest twice or more a year in a single plot. They consider multiple cropping an output diversification measure through the e$ cient use of the land restricted in area. However, crops must be carefully chosen and haphazard mixtures be avoided. For example, some villagers asserted that chili must not be cropped in a mixture with Fig.- . Land use in the,**, post-rainy season sorghum, as the intensive fertilization of chili leads (explored in Jan,**- ). to sorghum attracting pest insects to the chili. : Chili, : eucalyptus, : Wetland rice, : These facts indicate that villagers have their own Sorghum, : Cotton (incl. mixed-cropping), : logic behind the rejection of the modern technology Upland rice (incl. inter-cropping), : Other for some particular crops, and that villagers tradi- cropping patterns with red gram, : Others tionally have taken various measures to optimize (consisting only of sesame or leguminous crops), outputs. With the aim of providing more systemat- ῍ῌ: Irrigation wells dug on the riverbed, : ic viewpoints, the villagers’ perception of the rela- houses, : road, : tank bund, : tionship between crop and land must be examined. river or stream. Spatial distribution of cropping patterns and villagers’ soil classification The farmland map shows uneven spatial distribu- Villagers classify the soil into eight types. The tion of cropping patterns (Fig.- ). In the lower dominant types are called “nalla reagadi (black riverside zone, mixed cropping of sorghum and clay)” and “garubu neala (coarse soil)”. The for- green gram was common. Chili fields tend to clus- mer occupies the lower zones such as the riverside ter near irrigation wells dug in the riverbed. From plains along the Godawari tributary and the beds of ,**,to ,**. , sorghum and green gram were re- basins surrounded by small inselbergs, while the placed with chili in the greater part of this riverside latter is found in the upper zones. Soil profiles of zone (Fig.. ). In the upper zone, cropping patterns those two types showed morphological di# erences are much more diverse spatially and temporally. (Fig./ ). The “black clay” consists entirely of black Paddy fields are located in and around basins sur- (dark brown) silty clay loam without rock frag- rounded by small inselbergs. The discussion below ments and has vertical cracks in the dry season. indicates that farmers’ crop choice is closely related This soil type is known as vertisol in the USDA soil to soil moisture conditions of each plot. taxonomy (cf. USDA,+332 ). The “coarse soil” Tsunashima: Modern Agricultural Technology and the Koya People 233

Fig./ . Soil profiles of the two main soil types in Chalampalem. Soil type classification is according to Chalamp- alem villagers. These soil profiles were observed in June,**- (beginning of the rainy season). Soil type, crops planted in the year,**, , soil color and texture of each layer were as follows: Pit No.++ :Nalla reagadi (Black clay), Chili, Ap : +*YR -, / , SiCL, A ,+* : YR -, / , SiCL. Pit No.,++*-, :Nalla reagadi, Chili, Ap : YR / , SiCL, A,+*-,+*-+ : YR /ῌ YR / , SiCL. Pit No.- :Nalla reagadi, Sorghum῍ green gram, Ap+ : +* YR , / - , SiLῌῌ SiCL, A , : +* YR , . / / , , SiL SiCL. Pit No..+,/ :Nalla reagadi, Wet-land rice, Ap : . Y -,/ , SiLῌ SiCL, A ,,/,+ : .Y/, SiCL, A -,/-+ : .Y/, Fig.. . Land use in the,**. post-rainy season SiCL. (explored in Oct,**. ). Pit No./ :Gaubu neala (Coarse soil), Wet-land rice, : Chili (incl. nursery), : eucalyptus, : Wet- Ap:.Y/,SiL,/-,ῌῌSiCL, B +,/-, : .Y/,SiL SiCL, B:,,/-, .Y/,SiCL, B -,/-, : .Y/,SiCL, BC: +* YR - / land rice, : Sorghum, : Cotton (incl. mixed- -+*.., SCL, C: YR / , SLῌ LS. cropping), : Upland rice (incl. inter-cropping), Pit No.0 :Gaubu neala, Sorghum῍ sesame, Ap: : Other cropping patterns with red gram, : 1/.YR./.,SLL,Bt:. ,/ -/ῌῌ 21/ YR/,SiL - , SiCL, maize, : Others (consisting only of sesame or BC+1/ : . YR -- / , Sῌῌ SL, BC ,1/ : . YR .. / , S SL. leguminous crops),῎ : houses, : road, : Pit No.1 :Gaubu neala, Sesame, Ap: - . 1/ YR./, , / / SLῌῌ L, B+/ : YR .. / , L SiL, B ,/ : YR ,// . / , L ῌ tank bund, : river or stream. SiL, B-/ : YR./,L,B: ,// ./ YR/,L .. ῌ SiL. : Layers consist of sand (S), loamy sand (LS), sandy loam (SL), sandy clay loam (SCL) or loam varies in color and texture and has layers rich in (L). : Layers consist of silty loam (SiL) or rock fragments only/* cm below the ground sur- silty clay loam (SiCL). : Rock fragments. face. Indicating their size and abundance. : Very Villagers, needless to say, have never observed wide cracks (more than/ mm in width). Source: soil profiles. However, they have an insight into field work,**,ῌ ,**- . di# erence in texture of subsurface soil. For exam- ple, the soil pits No../ and (Fig. / ) are quite similar in texture and color of surface but com- capacity, which the cultivators can know only by pletely di# erent in those of subsurface; villagers experience. The following facts support this pre- distinguish between these two soils. This fact indi- sumption. cates that they do not classify the soil according to There is a clear di# erence in cropping patterns the appearance of surface. Rather, the criterion of adopted for those two soil types (Table. ). “Black the classification is presumed to be water-holding clay” lands are partly inundated and become too 234 J. Dev. Sus. Agr./, ( )

Table. . Relation of upland crops to the soil type

Proportion (ῌ ),ῌ Crop (variety) Cropping season BC : CS : Others Chili Post-rainy3. : / : + Black gram Post-rainy+** : * : * Sorghum (yellow)῍ Green gram Post-rainy0- : ,3 : 2 Sorghum (yellow)῍ Sesame Post-rainy* : +** : * Green gram Post-rainy* : +** : * Sesame Post-rainy, Rainy.30* : : Upland rice Rainy* : +** : * Cotton Rainy* : +** : * Sorghum (hill)῍ Little or Pearl millet Rainy* : +** : * Red gram+ῌ Rainy* : +** : * Source: sampled-household survey conducted in Aprῌ May,**- and all-household sur- vey conducted in Aprῌ May,**/ . +ῌ Including mixed- cropping with sesame, green gram or little millet. ,ῌ The values indicate average proportions in area cropped during the three cropping years from,**, to ,**. . BC: “Black clay” (nalla reagadi ), CS: “Coarse soil” ( garubu neala ). muddy for upland cropping in the rainy season. plot. Villagers change crops planted on each plot Thus, those lands are planted with upland crops in every year depending on rainfall at the beginning of the post-rainy season, especially for chili and mix- the cropping season or according to the subtle soil ture of sorghum and green gram, as well as the conditions, which are hardly reflected in the soil wetland rice with the longer growing period. On classification mentioned above. Some examples are the contrary, “coarse soil” is cultivable even in the given below: rainy season, but readily dries up in the dry season (a) Some farmers report that, generally speak- because of its smaller water holding capacity. Thus ing, if red gram is to be planted, where or those lands are used for various cropping patterns when soil water is relatively plentiful, in the rainy season, including wetland rice with the upland rice can be inter-cropped with red shorter growing period, and cultivated with highly gram, otherwise red gram should be sole- drought tolerant crops such as sesame also in the cropped. post-rainy season. Generally speaking, according (b) A household divided one of their plots of to the sample household survey, “black clay” is land into two sub-plots, one with harder considered to be more fertile than “coarse soil”. and one with softer soil. In the year ,**, For example, the mixed cropping of “yellow sor- (annual rainfall:33, mm) they planted ghum” (an early-maturing sorghum variety) and “hill sorghum” (another sorghum variety) green gram yields on “black clay” almost twice as with little millet on the former and sesame much as on “coarse soil” without any fertilization. on the latter. Stony fields with shallow The lowest yield of sorghum and green gram in surface soil are usually planted with “hill ,**.(Table , ) was due to the fact that most land sorghum” in this area. They wanted to of “black clay” were cultivated with chili, and grow upland rice with red gram on the sorghum and green gram were mostly grown on latter, however rainfall was not su$ cient. “coarse soil”. Thus, they waited until the post-rainy Crop choice by more subtle and changeable soil season and planted sesame, which is toler- moisture conditions ant of drought. In,**- (annual rainfall: As “coarse soil” varies in soil texture, there is+/,3 mm), they could plant upland rice variation in the range of crops suitable for each with red gram on the latter, because rainfall Tsunashima: Modern Agricultural Technology and the Koya People 235

was su$ cient. vested before the soil dries up. (c) A household mixed-cropped “hill sorghum” The spatial distribution for the four most popu- with pearl millet in,**, , but with little lar varieties in Chalampalem, namely, MTU- +*+* , millet in,**- , because pearl millet is more IR- 0. , MTU- 1*,3 and BPT- /,*. , indicates that tolerant to drought, but yields lower than “coarse soil” and “black clay” plots are ordinarily little millet. Villagers consider that the planted with early and late maturing varieties, re- good yield of pearl millet and little millet, spectively (Table/ ). Several villagers planted pad- when mixed-cropped with sorghum, are--* dy fields lined along a slope with two varieties of kg/ha and//* kg/ha, respectively. di # erent growing periods, namely, IR- 0. and MTU- (d) A household plants one of their lands with1*,3 , according to the di # erence in soil moisture cotton if they receive satisfactory rainfall at conditions along the slope. Villagers assume that it the beginning of the rainy season. If rain- will be risky to plant drought-prone “coarse soil” fall is not su$# cient, they put o their plots with late maturing varieties. This is why the sowing of any kinds of seeds until the begin- later maturing varieties are usually grown in a con- ning of the post-rainy season, when they dition with more e# ective inundation than the ear- sowed with “yellow sorghum” with green lier maturing varieties. The fluctuation in yields of gram after “the Godawari has come and the four main varieties during the survey period gone”. shows that the latest maturing variety, BPT-/,*. , (e) A household left one of their lands fallow in withered in,**, , when the severe drought hit, but ,**,, because rainfall was not su $ cient to gave highest yield in normal years (Table / ). This make soil soft and they could not plow it. fact also indicates that the trade-o# for higher yield But in,**- , however, they could plow and is longer growing period. This tradeo# is the most mixed-cropped maize, red gram and little important characteristic for the choice according to millet. hydrological conditions of each plot. These examples also imply that the villagers clas- Good example of interface between the modern and sify upland crops by their tolerance to drought. traditional technology Sesame and pearl millet are considered highly toler- If the above-mentioned indigenous perception of ant, while cotton and upland rice are rather vulner- tripartite association among fertilizers, weeds and able, and the others are moderately tolerant. Vil- pest insects applies to the wetland rice cultivation, lagers consider this drought tolerance in combina- the merit of this cropping system comes to light. tion with soil moisture conditions in each year for The longer the growing period is, the more manure yearly crop choice. or chemical fertilizers the variety receives. Howev- Choice of wetland rice variety er, the late maturing varieties such as MTU-1*,3 More than+2 wetland rice varieties have been and BPT- /,*. require little more frequent weeding grown in Chalampalem. The same explanation as and little more pesticide spray than early maturing o#- ered above also applies to the choice of wetland varieties (Table ). This implies that inundation rice variety. Villagers classify wetland rice varieties controls weeds and thus precludes infestation of into two categories by growing period: early and pest insect. Thus, wetland rice cultivation is the late maturing varieties are called “pinna vari (small best use of the basin topography into which rainwa- rice crop)” and “pedda pyoori (large rice crop)”, ter flows. It can be said that the wetland rice cropp- respectively. A number of villagers have forgotten ing system is a good example of interface between the names of the varieties they have been using, the modern and traditional technology. though remembering the growing periods of those varieties. Since the wetland rice cultivation also New Trend in Land Use depends on rainfall or small-scale tank irrigation, Development of land lease market there are only two measures to reduce the risk of Simultaneously with the increase in cropped area drought. The first one is to cultivate plots with high of chili and cotton with an increased number of water holding capacity, and the second one is to cultivators, land lease contracts related to those adopt early maturing varieties that can be har- cash crops have also become common (Table0 ). 236 J. Dev. Sus. Agr./, ( )

Table/ . Relation of growing period to soil type and yield

ῌ ,ῌ Yearly yield (kg/ha) GrowingProportion ( ) [Yearly rainfall (mm)] Expected Variety period yield (days)+ῌ ,**, ,**- ,**. (kg/ha)-ῌ BC : CS : Others [][][]33- +/-* ++2. MTU-+*+* 3* . 30 * 33 +1,0 //3 n.a. IR-0. 3* ++ 21 , /- ,+,. -.1 -+.* MTU-1*,3 +*/ /1 -3 . +,, -1/- +/21 .-/1 BPT-/,*. +,* 13 ,+ * * -2*+ ,--. /2*/ Source: sampled-household survey conducted in May - June,**- , all-household survey con- ducted in Aprῌῌ May,**- and Apr May ,**/ . +ῌ According to villagers. Days required from transplanting to harvesting. ,ῌ The values indicate average proportions in area cropped during the three cropping years from,**, to ,**. . BC: “Black clay” (nalla reagadi ), CS: “Coarse soil” ( garubu neala ). -ῌ Expected as good yield in plots of the sampled households who actually adopted the variety is averaged.

These are rent-fixed single-year contracts, called after running deficit in chili cultivation. “kaulu vavyasaayam (lease farming)”. In the group Significance of rent-fixed land lease system interview, villagers explained that, generally speak- Formerly, the tenants used to be mostly non- ing, in the case of running a deficit or need to raise tribal farmers in Nandigama, who are said to be money, they provisionally lease out their lands and more a% uent than Chalamaplem villagers. How- work under the employ of the tenants or other ever, during the survey period, the case of land cultivators. The benefit of land lease for chili cul- lease between Chalamapalem villagers has in- tivation is the payment of an assured amount in creased in number (Table0 ). Even some landless cash at the beginning of the cropping season. Land households betted on leasing in pieces of land, lease is the best possible use of the land which the although the tenants have to take all the risk in the owners cannot cultivate. The advantage of wage rent-fixed land lease system. labor is also income in an assured amount. Since The result in,**. shows a striking di # erence cash cropping is accompanied by the risk of a between the households who leased in some plots sudden drop in selling prices, the land rent itself is and those cultivated only their own land (Table1 ). sometimes a more reliable source of income than The both spent about the same amount of money in profits from chili cultivation for the landowners. the total. However, the former had better yield Actually, the selling price of chili decreased from than the latter at a significant margin. Moreover, ,**,to ,**. (Table , ). Especially, the sharp de- the former applied more chemical fertilizer and less cline in the year,**. was much greater than villag- pesticide than the latter. This result suggests that ers had anticipated. As a result, the average net re- the plots leased in and out are more suitable for venue of chili cultivation was Rs.῍/ , -0* /ha chili cultivation than the other chili fields, or that (Tsunashima,,*+* ). In the meantime, the land the tenants can achieve higher yields in their own rent increased (Table0 ) and the minimum daily plots than the other chili cultivators. In either case, wage also rose from Rs.,/ to Rs. -/ . this result means that those plots were worth leas- The land lease for chili cultivation must also be ing to the tenants despite the apparent disadvantage compared with mixed-cropping of sorghum and in the land lease contracts. green gram in the fields of “black clay”. However, This land lease system has a function of redis- the choice between land lease and cultivation of tributing cash income within a village community. sorghum and green gram seemed to be a delicate Considering there have been many cases where the matter concerning diet; a few households returned tenant is more a% uent than the landowner, it can back to this traditional cropping system in,**/ also be said that this land lease system have some- Tsunashima: Modern Agricultural Technology and the Koya People 237

Table0 . Land leased by Chalampalem villagers+ῌ

System/Crop ,**, ,**- ,**. No. of Cases .2 /2 1. Land lease/Total Area (ha) -*4- .*4. ./4- No. ,2 .0 .3 Land lease/Chili Area +342 ,24/ ,34/ Rent (Rs/ha),ῌ /.,. 0/2* 0/,+ No. 3,/,1 Between Chalampalem villagers Area .4. +,4/ +-4+ No. +0 ,* +1 To non-tribal farmers in Nandigama Area +.4- +/4, +-40 No. +*+ From non-tribal farmers in Nandigama Area *40 *40 No. /0++ Land lease/Cotton Area -4* 04+ 34- Rent +0+- ,-22 ,*/* No. +*. Between Chalampalem villagers Area *42 *4* ,4, No. -01 To non-tribal farmers in Nandigama Area ,4* 04+ 14- No. *** From non-tribal farmers in Nandigama Area No. +* +/ ++ Land lease/Wetland rice Area .41 /42 /4. Rent -+.1 ,20* .0*3 No. 0+*3 Between Chalampalem villagers Area ,41 -4/ -41 No. *+, To non-tribal farmers in Nandigama Area *4. +40 No. -,* From non-tribal farmers in Nandigama Area +4/ *42 No. .*, Land lease/Sorghum (῍ Green gram) Area ,4. *41 Rent ,.1* 0*1 No. +*+--ῌῌ Land lease/Green gram (῍ Sesame) Area *4. *4. Rent 0+1/ 0+1/ No. -** Share cropping/Total Area ,4* Source: all-household survey conducted in Aprῌῌ May,**- and Apr May ,**/ . +ῌ The lands that Chalampalem villagers either leased in and out are both taken into account. ,ῌ Rs: rupees. -ῌ Cultivated in a land parcel suitable for chili. 238 J. Dev. Sus. Agr./, ( )

Table1 . Comparison in the results of chili cultivation in the year,**. between the households with some chili fields leased and those with no chili fields leased

Chili Costs for Total Leasing in ofNo. cultivation chamical Costs for production Yield Gross land for chiliof hh scale fertilizers pesticides cost (kg/ha) return cultivation(ha/hh) (Rs/ha) (Rs/ha) (Rs/ha) (Rs/ha) Some plots,, *41/+ῌ a ++0/+ a +,,,+ a /1-2- a -+00 a 1*1+. a No plot.3 *4/. a 212+ b +/-2- a //332 a ,.32 b /0200 a Source: all-household survey conducted in Mayῌ June,**- and April ,**/ . +ῌ Area of the land leased in was*.0 . ha/hh on average. Note: means followed by the same letters are not significantly di#/ erent atῌ level by t-test.

times realized “trickle-down” through the introduc- done before “sowing seeds”. tion of the modern technology. Incidentally, in the As reviewed in the introduction, past research study area, landless households in distress can use a had an expectation that the ethnic minorities would quasi land tenancy system called “pottulu vavyasa- someday adopt mainstream agricultural technolo- ayam (joint farming)”, in which landowners and gies as a complete package. The mainstream agri- other households cultivate together a part of or the cultural development, which started with the green entire farm of the landowner and share both pro- revolution, intended to enable farmers to sell their duction costs and produce, without any payment of surplus produce (Warren,,**, ). However, against land rents. This system functions as a safety net for these intentions, the food production of the Koya poor landless households, and sometimes as a risk- people produces little surplus for sale. sharing measure adopted for chili cultivation. In order to earn more cash income, villagers have Conclusion introduced new cash cropping systems instead. The high-input system for cash cropping of chili is cost- The present study illustrated that, because of ly, and also exposed to the mainstream market geographical constraints in installing irrigation economy. Thus, chili cultivation is accompanied by facilities, it is still impossible for the Koya people to the risk of a sudden decline in the selling prices of replace their entire farming system with main- the products. Consequently, land lease contracts stream technologies. They have only adopted the between risk-averse landowners and risk-taking whole high-yielding package for chili cash cropp- tenants have become common. ing. They do not entirely depend on the high- There is an apparent contradiction between the yielding packages of the technol- general economic theory about the market econo- ogies for their food requirements even after intro- my and the fact that, while the selling price de- ducing the chili cash cropping system. Their choices creased, the land rent and labor wages increased. from a high-yielding package of cultivation tech- This contradiction suggests that some local non- nologies are based on their agroecological percep- market institutions regulate land rents and labor tion. wages independently of the chili wholesale market. Elder villagers explained the change in their Although the extent to which the local decision is farming practices as follows: formerly cultivators independent of the outside market awaits further would only sow seeds and wait for harvest, but investigation, it is clear that the Koya people view today they use lift irrigation, chemical fertilizers land tenancy in a di# erent light than past research. and pesticides. The traditional cropping systems Generally speaking, the past research discussed the consist of knowledge intensive farm tasks such as risk reduction measures of tenants (Anderson, decisions on fertilization and choice of crop/varie-,**- ; Ramaswamiet al., ,**- ). On the contrary, ty. It is this idea that they explained as “just sowing the present study suggests that the rent-fixed land seeds and waiting for harvest”. These tasks are apt tenancy can be viewed as a strategy of landowners to be invisible to outside professionals because it is to stabilize their income. Importantly, this system Tsunashima: Modern Agricultural Technology and the Koya People 239 has involved the relationship between the Koya Indian Agriculture: Cropping Pattern Changes and $ people and non-tribal people from Nandigama. E ciency Gains in Semi-Arid Tropics. Oxford Univer- sity Press, New Delhi. This study implies that landowners and tenants Guilmoto, C.Z.,,**, . Irrigation and the great Indian rural discern the land which is the more suitable for chili database: vignettes from South India. Economic and cultivation. The agroecological perception of the Political Weekly-1 ( +- ), +,,-ῌ +,,2 . Koya people turns to be valuable for redistribution Hansda, S.,+33. . Problems in agricultural development in of cash income through land lease. Research after tribal areas. In: Thakur, D., Thakur, D.N. (Eds.), Tribal Agriculture and : Tribal Life the green revolution discussed whether the intro- in India. Deep & Deep Publications, New Delhi, pp. duction of the modern technology widens the dis- ,2+ῌ ,21. parity between the rich and the poor (cf. Freebairn, Iyengar, S.,+33. . Issues in agricultural development in a +33/). The land lease system of the Koya people tribal area - a Case Study of Panchmahals District. In: may give us useful hints on designing an alternative Thakur, D., Thakur, D.N. (Eds.), Tribal Agriculture and Animal Husbandry: Tribal Life in India. Deep & to the mainstream agricultural development so that Deep Publications, New Delhi, pp.,22ῌ -,/ . the disparity can be reduced. Kornel, D.,,**0 . Tribal Crop-Livestock Systems in South- Considering that there are still many villages East India. Manohar, New Delhi. Mehta, A.K. and Shah, A.,,**- . Chronic poverty in India: unblessed with water resources for lift irrigation -+ ,**3 incidence, causes and policies. World Development (Tsunashima, ), the traditional cultivation (),- .3+ῌ /++ . method for upland crops is still vital for the major- Ramaiah, P.,+32+ . Tribal Economy of India: A Case Study ity of the Koya people. The future agricultural of Koyas of Andhra Pradesh. Light and Life Publish- development for the Koya people should be built on ers, New Delhi. ,**- an accurate understanding of local institutions as Ramaswami, B., Ravi, S. and Chopra, S.D., . Risk Management in Agriculture. Discussion Paper*-ῌ *2 . well as the Koya’s knowledge of their agroeco- Indian Statistical Institute, Delhi. system. Shah, A. and Guru, B.,,**- . Poverty in Remote Rural Areas in India: A Review of Evidence and Issues, Acknowledgements Working Paper No.+-3 . Gujarat institute of Develop- ment Research, Ahmadabad. I am thankful to all the villagers in Chalampalem, Subramanyam, V.,,**- . , Technology and Re- and all the members of local NGOs, KRUTHI. sources Management among the Tribes of Eastern Also I am deeply grateful to A. Murali (KRUTHI), Ghats: An Anthropological Study. http://www.sasnet. Walter Mendoza (Godawari Basin Action Group) lu.se/tribalsubraman.doc +33. and Kazuo Ando (Center for Southeast Asian Thakur, D. and Thakur, D.N., . Introduction: land and the tribe. In Thakur, D., Thakur, D.N. (Eds.), Tribal Studies, Kyoto University) for their support. I Agriculture and Animal Husbandry: Tribal Life in would like to express my gratitude to S. P. Wani, V. India. Deep & Deep Publications, New Delhi,+,/ῌ . Balaji and S. V. Prasad Rao for accepting me as a Tsunashima, H.,,**3 . Significance of informal trade in Research Scholar in International Crops Research non-timber forest products for the Koya people, an ethnic minority in Andhra Pradesh, India. Tropical Institute for Semi-Arid Tropics, Patancheru. Tim Agriculture and Development/- ( - ), 1.ῌ 2+ . Olive and Steven Ivings took the trouble to read the Tsunashima, H.,,*+* . Extension of commercial agriculture manuscript. The,+ st Century COE Program, in a “tribal” ethnic minority area in Andhra Pradesh, Graduate School of Asian and African Area India: The case of chili cash cropping among the Koya Studies, Kyoto University subsidized this research. people. Journal of Agricultural Development Studies ,*(), - 3ῌ +2 . Referencres United States Department of Agriculture and Natural Re- source Conservation Service,+332 . Keys to Soil Taxon- Anderson, J.R.,,**- . Risk in rural development: challenges omy,2 th ed. United States Department of Agriculture for managers and policy makers. Agricultural Systems and Conservation Service, Washing- 1/,. +0+ῌ +31 ton D.C. Freebairn, D.K.,+33/ . Did the green revolution concen- Warren, P.,,**, . Livelihoods Diversification and Enter- trate incomes? A quantitative study of research reports. prise Development: An Initial Exploration of Concepts World Development,- ( , ), ,0/ῌ ,13 . and Issues, LSP Working Paper. . FAO. Gulati, A. and Kelley, T.,+333 . Trade Liberalization and