PRODUCTION PERFORMANCE OF MAIZE IN INDIA : APPROACHING AN INFLECTION POINT

In. J. Agricul. Sa. Sci., Vol. 10, No. 1, pp. 241-248, 2014 ISSN : 0973-1903 ORIGINAL ARTICLE PRODUCTION PERORMANCE O MAIZE IN INDIA : APPROACHING AN INLECTION POINT Ranji Kumar*, K. Srinivas, Naveen Kumar Boiroju and Pravin C. Gedam Naional Academy of Agriculural Research Managemen, Rajendranagar, Hyderabad - 500 030, India. E-mail: ranjikumar@naarm.erne.in Absrac : Maize is he hird mos imporan crop afer rice and whea in India in erms of is area coverage and conribuion o oal food grains producion. I is grown in almos all agro-ecological regions, conribuing abou 22 million onnes of grain producion from abou 9 million hecares of land. The producion paern as well as is consumpion paern has dramaically changed in India in he recen pas. However, he average crop yield of 2.5 /ha is sill very low as compared o he oher Asian counries like Bangladesh (6.8 /ha), China (5.7 /ha), Indonesia (4.6 /ha) and Pakisan (3.8 /ha) in 2012 (AOSTAT). This paper has examined he growh and insabiliy in maize producion in he major disrics of major maize growing Saes in he counry. An aemp has also been made o forecas is producion in near and mid-erm under differen scenarios. The resuls showed ha more han 60 per cen of maize area is observed o be having maize yield less han 2 /ha. A he same ime, huge variabiliy in maize yield beween and wihin he maize growing Saes was also found. However, he maize yield is esimaed o increase lile more han 3 /ha by he year 2020, if he curren policy and macro-economic environmen coninues. I is also expeced ha he demand for maize in domesic as well as in inernaional marke will grow faser han he producion growh, creaing very good poenial for Indian maize in he near fuure. Therefore, o mee he fuure growing demand of maize in domesic as well as inernaional marke, i is desired o have differen growh pahway han he exising rend. To realize ha kind of inflecion poin, India has o gear up is logisics and supply chain o handle he near-double size of he maize oupu in coming years. Key words : Maize, orecasing, Growh, Insabiliy, Box-Jenkins mehodology and Arificial Neural Neworks. 1. Inroducion Maize (Zea mays L) is a cereal grain, also known as Queen of Cereals due o is diverse usages. In India, i is culivaed in mos of he Saes hroughou all he seasons. Depending on he regions and socio-economic condiions of he populaion, he maize-grain is used for various purposes including food, feed, fodder, green cobs, swee corn, baby corn, popcorn, sarch and several indusrial producs. According o he fourh advance esimaes of he Minisry of Agriculure, Governmen of India, maize in India occupied abou 8.7 million hecares (M ha) of he area and produced abou 22.2 million onnes (M) of maize grain during 2012-13. I is abou 15 per cen and 5 per cen o oal maize-area, while 8 per cen and 2.4 per cen o oal producion in Asia and he world, respecively [AOSTAT (2013)]. In he pas wo decades, producion performance in India has been significan as compared o he previous periods. The oal maize producion in he counry has doubled wihin 15 years from abou 10 M in 1992-93 o 20 M in 2008-09. However, his growh can be equally aribued o area expansion under he crop as well as improvemen in he yield, as boh has increased by around 40 per cen in he same period. During his period, he counry has winessed a econic shif in maize growing regions. Till 1990s, Bihar, Madhya Pradesh, Rajashan and Uar Pradesh were he major maize-producing Saes, bu he souhern Saes especially Andhra Pradesh and Karnaaka became he predominan maize-growing saes from he las wo decades [Gulai and Dixon (2008)]. More recenly, Maharashra and Tamil Nadu are also emerging as very imporan maize growing Saes. The new erriory of he crop has also pushed he crop yield up significanly due o he clemen weaher condiion in he region. Alhough, he improvemen in average crop yield (2.55 /ha in 2012-13) has no been very encouraging, which is far lower han ha realized in many oher Asian counries like Bangladesh (6.8 /ha), China (5.7 /ha), Indonesia (4.6 /ha) and Pakisan (3.8 /ha) in 2012. The maize producion in India has been largely driven by he growing demand from he feed indusry and various indusrial usage. According o exising lieraures, he uilizaion paern of maize a presen includes 51 per cen *Auhor for correspondence. Received Sepember 24, 2013 Revised ebruary 08, 2014 Acceped May 17, 2014

242 Ranji Kumar e al. as poulry feed, 20-25 per cen as human food, 10-12 per cen as cale feed, abou 10-12 per cen going owards indusrial processing like sarch and brewery and 1 per cen as seed [DMR (2012) and USDA (2013)]. As he demand for maize is increasing due o is muliple uses, i is imporan o undersand he exising maize siuaion, so as o plan he fuure based on he pas and presen siuaion. On he oher hand, i should also be kep in consideraion ha he spaio-emporal variaions in projeced changes in emperaure and rainfall are likely o lead o differenial impacs on maize yield in he differen regions in India [Kaarkandi e al. (2010)]. In his paper, he curren siuaion of maize producion in India has been analyzed and forecasing of area, producion and yield (APY) of maize in India has been done under differen scenarios. The findings are expeced o provide he guidelines o differen sakeholders in he maize secor o mee he changing demand in he fuure. 2. Daa and Mehodology or he sudy, he high frequency daa a disriclevel on area, producion and yield of maize was colleced for he period 1986-87 o 2012-13 from he Direcorae of Economics and Saisics, Deparmen of Agriculure and Cooperaion, Minisry of Agriculure, Governmen of India. The growh performance of area, producion and yield of maize was analyzed by esimaing compound annual growh raes (CAGR) by fiing a semi-log rend equaion of he following form [Barle (1993)] ln Y = a + b where Y defines he ime series daa of APY of maize, is he rend, a is he consan coefficien and b is he ln (1 + CAGR). Thus, CAGR = {anilog (b) 1}*100. Insabiliy is one of he vial decision parameers in developmen dynamics and more in he conex of varying agriculural area and yield. The high growh accompanied by low levels of insabiliy in yield for any crop is desired for susainable developmen of agriculure [Tripahi and Prasad (2009)]. The insabiliy in maize producion is measured by he Cuddy-Della Valle index (CDI). The index correcs he coefficien of variaion (CV), which overesimaes he level of insabiliy in ime-series daa characerized by long erm rends [Cuddy-Della Valle (1978)]. I is esimaed by CDI = CV* (1 adj R 2 ) 0.5, where CV is he Coefficien of variaion and R is he coefficien of deerminaion. Three forecasing models namely, auoregressive inegraed moving average (ARIMA) model, growh model and arificial neural neworks model were considered for he predicion of area, yield and producion of maize for he years 2012 o 2020, based on he daa available for he period 1986-2011. The ARIMA echnique is based on he principle ha a saionary process can ofen be parsimoniously represened by a mixure of auoregressive and moving average models. Even a non-saionary series afer proper differencing, can be reaed as a saionary series [Box e al. (1994)]. In case of growh model, Z expba bg e is esimaed, where Z denoes he forecas value; represens he ime (year) and e represens he error. The consans of he growh model are deermined using leas squares mehod. Arificial neural neworks (ANN) are inspired by biological sysems, paricularly by research ino he human brain. Currenly, ANNs are being used for a wide variey of asks in many differen fields of business, indusry and science. One major applicaion of ANNs is forecasing, which is used as an alernaive ool for boh forecasing researchers and praciioners. Haykin (1999), Zhang e al. (1998) and Boiroju (2012) presens he neural neworks mehodology and is applicaion in ime series forecasing. In feed forward neural neworks, he relaionship beween he oupu Z and he inpus Z -1, Z -2,, Z -p is as follows: Z 0 j f ij i oj e in which 0 and oj denoe he weighs of he connecion beween he consan inpu (bias) and he oupu and beween he bias and hidden nodes, respecively and p is he number of inpu nodes, q is he number of hidden nodes, while { j, j = 1, 2,, q} is a vecor of weighs from he hidden nodes o oupu nodes and { ij, i = 1, 2,, p; j = 1, 2,, q} are weighs from he inpu nodes o hidden nodes. f denoes he ransfer funcion used in he hidden layer. Transfer funcions such as he logisic or hyperbolic angen funcions are commonly used for ime series daa; as hey are non-linear and coninuously differeniable, which are desirable properies for nework learning [Kasra and Boyd (1996)]. eed forward neural neworks is used wih hyperbolic angen funcion as an acivaion funcion under wo hidden unis in a single hidden layer and he ime variable aken as he inpu for he nework o predic he fuure APY of maize. The above-menioned hree ime series mehods were used o predic he area and yield of maize in major Saes. Among hese, he model was seleced which has he minimum forecasing error. In his sudy, i was observed ha he ANN models performed beer han he growh and ARIMA models. Therefore, he forecass of maize q j= 1 p i= 1 I K J

Producion Performance of Maize in India : Approaching an Inflecion Poin 243 area and yield were esimaed using ANN models represened by scenario-i or Business As Usual (BAU). The producion of maize was aken as he produc of he area and yield. Two more alernae scenarios were also developed. Under scenario-ii, i was assumed ha he area under he crop would expand according o is pas rend or BAU-way, bu adopion of hybrid may increase up o 75 per cen of he maize area by 2020. A he same ime, he yield of hybrid maize may grow o reach up o 5 /ha by 2020. Under Scenario- III, i is assumed ha here may no be any furher growh in maize area from he curren level of 8.7 M ha, however all effors may be made o cover 90 per cen of he maize area by hybrids wih an average yield growing o ouch 5 /ha by 2020. 3. Maize Trends in India In his secion, he resuls of pas performances of he area, producion and yield of maize for differen Saes are presened. rom perusal of he findings, i was observed ha four Saes namely Karnaaka, Rajashan, Andhra Pradesh and Madhya Pradesh ogeher consiued abou 48% of he oal maize area in he counry. As compared o TE 2000-01, he percenage share of Karnaaka, Andhra Pradesh, Maharashra and Tamil Nadu o he oal maize area in he counry has increased significanly in TE 2010-11, while i has declined in Rajashan, Madhya Pradesh, Uar Pradesh, Bihar, Himachal Pradesh and Punjab (ig. 1). The area under he crop in Tamil Nadu, Maharashra and Karnaaka has escalaed by 3.47, 2.54 and 2.07 imes, respecively during his period. Conrary o i, he radiional maize growing Saes like Bihar, Himachal Pradesh, Madhya Pradesh and Uar Pradesh have winessed a decline in he crop area during he same period. I is imporan o noe ha he saes where maize area sared conracing are radiional maize bel. The maize-growers in hese Saes prefer o grow open pollinaed varieies (OPVs) or composies, mainly for domesic consumpion. Besides, i is mainly grown in Kharif season as rainfed crops wih low inpus. The yield in Odisha, Himachal Pradesh, Rajashan, Uar Pradesh, Chhaisgarh, Gujara and Madhya Pradesh was below he naional average yield (ig. 2). On he oher hand, in Andhra Pradesh, Tamil Nadu, Karnaaka and Maharashra, here was significan jump in he crop yield due o beer hybrid adopion, as well as relaively high inpu use [ASG (2012)]. Andhra Pradesh and Madhya Pradesh have almos he same area, bu hese wo Saes differed significanly in he yield level. Similarly, Uar Pradesh and Maharashra has he same area, bu he yield is 1.48 /ha and 2.50 /ha, respecively, while he maize area in Himachal Pradesh and Tamil Nadu are equal, bu hese wo Saes differed significanly in he yield a 2.13 /ha and 4.44 /ha, respecively in TE 2010-11. In erms of producion, four Saes Karnaaka, Andhra Pradesh, Maharashra and Tamil Nadu ogeher conribued abou 55 per cen of oal maize producion in India. During he pas en years, he highes change was observed in Andhra Pradesh, Maharashra and Tamil Nadu (ig. 3). A significan increase in he maize area and producion during his period has happened mainly due o he inroducion of single crossed hybrids and implemenaion of Governmen of India sponsored Inegraed Scheme of Oilseeds, Pulses, Oilpalm and Maize (ISOPOM), as well as shif in growing season from Kharif o Rabi in many Saes [Dass e al. (2010) and DMR (2012)]. or he popularizaion of hybrid maize in he counry, several Sae governmens also sared promoional programmes in public-privae-parnership mode o launch he projecs like Projec Sunshine in Gujara, he Golden Rays Projec in Rajashan, Projec Golden Days in Odisha, Makka Vikas Pariyojana in Madhya Pradesh and he Projec Rainbow in Jammu & Kashmir. 4. Spread and Growh Performance of Maize The pas growh of maize in several Saes has pushed oher crops o he margin in India. During he pas five years, here was a negaive growh in he area for sunflower, bajra, ragi, groundnu, jowar, small milles and coon o some exen in Andhra Pradesh, Maharashra and Karnaaka a he cos of maize. Iniially, maize had replaced rice and whea in Bihar, Madhya Pradesh and Uar Pradesh, however laer crops sared regaining is los ground. According o he maize producion daa for he year 2011-12, all he maize-growing Saes wih more han one per cen of he oal maize area in India can be caegorized as High maize produciviy saes: (yield: > 4 /ha) Medium maize produciviy saes: (yield: 2-4 /ha) Low maize produciviy saes: (yield: < 2 /ha) Andhra Pradesh and Tamil Nadu, ogeher consiue abou 13 per cen of he oal maize area in he counry. Bihar, Himachal Pradesh, Karnaaka, Maharashra, Odisha, Punjab and Wes Bengal ogeher consiue abou 40.15 per cen of he oal maize area in he counry. Chhaisgarh, Gujara, Jammu & Kashmir, Jharkhand, Madhya Pradesh, Rajashan and Uar Pradesh ogeher represening 43.78 per cen of he oal maize area in he counry. In erms of maize yield, he op en disrics were: Gunur (9.7 /ha), Krishna (7.8 /ha), Nellore (7.5 /ha),

244 Ranji Kumar e al. Source: DES, Minisry of Agriculure, Gov. of India ig. 1 : Change in maize area in major growing saes in India. Source: DES, Minisry of Agriculure, Gov. of India ig. 2 : Change in maize yield in major saes in India. Source: DES, Minisry of Agriculure, Gov. of India ig. 3 : Change in maize producion in major growing saes in India. Wes Godavari (7.2 /ha), Prakasam (7.2 /ha), Cuddapah (7.2 /ha), Eas Godavari (6.9 /ha) and Srikakulum (6.28 /ha) from Andhra Pradesh and Erode (7.2 /ha) and Pudukkoai (6.5 /ha) from Tamil Nadu. More imporanly, major chunk of maize area i.e. abou 44 per cen of maize area, spread across he res of he Saes harvesed less han 2 /ha of maize yield, which invies special aenion of researchers and policy makers. Mos of he maize-growing disrics in Gujara, Madhya Pradesh, Odisha, Rajashan, Uar Pradesh, Jharkhand and Norh-Easern saes had yields in he range of 1-2 /ha. rom disric-level analysis, i has been observed ha only 13 disrics spread across five Saes had more han 100 housand hecares area under maize crop (ig. 4). These were Karimnagar and Mahaboobnagar in Andhra Pradesh; Dahod and Panchmahals in Gujara; Belgaum, Davangere and Haveri in Karnaaka; Aurangabad and Nasik in Maharashra and Banswara, Bhilwara, Chiorgarh and Udaipur in Rajashan. The highes maize area in India was observed in he Bhilwara disric of Rajashan. Ou of 640 disrics in India, maize was culivaed in 524 disrics, ou of which 408 disrics in 25 Saes had a leas 500 ha maize area, while 168 disrics had more han 10 housand ha under maize culivaion. Maize produciviy also varies widely across he disrics in hese Saes of India. There were 76 disrics which harvesed maize grain more han 3 /ha, culivaing maize area in he range of less han 500 ha o more han 50,000 ha. In conras, 340 disrics had maize yield less han 2 /ha. These disrics had maize area ranging from less han 500 ha o 10,000 ha. There were hree disrics in Rajashan having more han 25,000 ha of area under maize culivaion, bu harvesed less han 1 /ha of maize grain. The reasons behind such a low yield in hese Saes are: maize is grown as a rainfed crop and secondly, i is used mainly as a food crop, for which radiional or composie varieies are preferred giving poor yields. In his conex, i is imporan o noe ha many rainfed disrics have large maize-area bu wih poor yield. The growh in area, producion and yield of maize in major Saes in India are presened in Table 1. I can be observed ha he area under he crop and is yield have grown by abou 3 per cen annually in he pas decade in India, hough he area growh has declined in recen years, while ha of yield has acceleraed o 4 per cen. Among he major Saes, he pace of area expansion under maize has been faser in Tamil Nadu, Maharashra, Karnaaka and Andhra Pradesh during he period 1991-2011. Similarly, he growh in maize yield has been fases in Tamil Nadu followed by Odisha, Andhra Pradesh, Punjab, Maharashra and Karnaaka during he period 2001-2011. However, opposie o i, negaive growh in area as well yield was recorded in Bihar and Himachal Pradesh during he pas five years. Thus, new producion regions like Andhra Pradesh, Karnaaka, Maharashra, Odisha and Tamil Nadu have shown remendous poenial wih very high growh boh in acreage and produciviy of maize in recen years.

Source: Auhor.s calculaions ig. 4 : Disribuion of 524 maize-growing disrics in India according o maize yield and area under he crop, TE 2010-11. Source: Auhors. esimaion ig. 5 : orecass of APY of maize in India under scenario-i. Source: Auhors esimaion Producion Performance of Maize in India : Approaching an Inflecion Poin 245 ig. 6 : Expeced share of major saes in oal maize producion in India (2015-2020). In lae 1990s, use of single cross hybrids and a significan shif from Kharif o Rabi season for maize culivaion in several Saes came ino being rapidly in mos of he Saes. The improved growh in he APY of maize could happen due o he new paradigm shif in he seed policy, supporive price policy, acive paricipaion of privae seed companies in developing hybrid seed marke and huge demand racion from poulry secor as feed. The growh in producion during his period was well suppored by an increase in acreage apar from an increase in he yield. The growh and insabiliy of maize yield in major maize growing Saes in India are presened in Table 2. In India, maize area is growing only in few Saes, like Andhra Pradesh, Karnaaka, Maharashra and Tamil Nadu. In oher Saes, i eiher seized o expand or sared declining. As far as he growh and insabiliy in he yield are concerned, Tamil Nadu and Odisha have noiced high growh in yield wih high and low insabiliy, respecively. Andhra Pradesh, Karnaaka and Chhaisgarh have grasped medium growh in yield wih medium insabiliy while Maharashra and Punjab have accrued medium growh in yield wih low insabiliy. Rajashan alone has more han 13 per cen area, bu had slow growh in he yield wih high insabiliy. This is mainly due o he fac ha maize is grown in Kharif season in he sae wih low inpus and almos no irrigaion. More disappoining siuaions appeared from Gujara, Madhya Pradesh and Bihar, where he maize has winessed negaive growh in yield wih high, medium and low insabiliy, respecively. 5. Oulook of Maize As in he 2013 Kharif season, maize planing has se he pace in India and if weaher condiions remain clemen, he rend indicaes ha maize-growers inend o plan he larges area ever. The forecas of maize area, producion and produciviy was done for all he major growing Saes on he basis of pas experiences of echnological and policy changes implied in he previous observaions. The resuls assume ha if he pace of change in policy, echnology adopion and inpu use would remain same. These forecass are subjec o saisical errors and hey are indicaive of he oulook of maize in major maize-producing Saes. The maize area and yield in he counry are denoed by X and Y, respecively. The ANN predicion model for he maize area and yield are given by a ab 0 1 1 2 2g and Y y yb h 0 1 1 2 h 2 X h h g

246 Ranji Kumar e al. respecively, where h Tanh h Tanh 2 21 22 II KJ KJ 1 11 21 II KJ KJ for = 1986, 1987,, 2020. The mean and sandard deviaions of he variables are represened by µ and, respecively. The variable is sandardized wih linear-norm mehod o reduce he error. The parameers of he predicion model for he maizearea in he counry are approximaed o hree decimal places and are given by µ = 1999, = 8.5, µ a = 6852.497, a = 1038.045, 11 = -0.008, 21 = 0.381, 12 = -0.8, 22 = 1.157, 0 = 0.525, 1 = 0.452, and 2 = 1.36. Similarly, he parameer values for he predicion model of he maize-yield in he counry are µ = 1999, = 7.415, µ y = 1822.886, y = 391.198, 11 = -1.398, 21 = 0.828, 12 = -0.679, 22 = -0.348, 0 = 0.296, 1 = 1.854 and 2 = -1.98. The maize producion in India was esimaed by aking he produc of he area and yield. The forecass of he area, producion and yield presened in he ig. 5, reveal ha he area under maize may furher increase in fuure also and may reach o abou 9 M ha. Similarly, he producion is also going o boos up wih he addiion of more han 4 M of maize grain. However, he yield may be increasing slowly and likely o improve marginally from curren level of 2.55 /ha o 3.17 /ha in 2020-21. The increase in he maize producion in he fuure is expeced o come from Karnaaka, Maharashra and Tamil Nadu Saes. The maize area is expeced o furher grow in Andhra Pradesh, Maharashra, Karnaaka and Tamil Nadu (ig. 6). Karnaaka is expeced o mainain is prime posiion in he near fuure also, wih he oal maize producion of 5.7 M by 2020-21. In Maharashra, maize is expeced o ouperform on accoun of a huge expansion in he maize area o 1.2 Mha and hereby producion will rise o 4 M while produciviy will increase o 3.4 /ha by 2020-21. Tamil Nadu is one of he fases emerging maize-producing Saes in India because of growing poulry and feed indusries. The area, producion and produciviy of maize in he Sae would be 0.32 Mha, 2.0 M and 6.25 /ha, respecively. Bihar, Madhya Pradesh and Uar Pradesh are he radiional maize-growing Saes, where opporuniy for area expansion under maize is very limied and he gain in is producion is expeced only from produciviy gain due o improvemen in adopion of hybrid seeds and beer seed replacemen rae. In Punjab, even afer all he effors by he governmen for crop diversificaion, maize may sruggle o gain exra acreage, however, gain in maize produciviy would be greaes among all he Saes. Oher Saes like Odisha, Gujara, Chhaisgarh and Rajashan have also mixed resuls, where maize area or yield or boh may, increase in he fuure wih differen paces. During he pas eigh years (2005-2012), he average annual incremen in maize area was 140 housand hecares and in maize yield was 77 kg/ha. However, he forecass show an average annual incremen in maize area of abou 25 housand hecares and in maize yield of abou 64 kg/ha in he nex eigh years (2013-2020). Thus, he growh in maize producion is expeced o be more yield-based han he area-based. Currenly, i is esimaed ha only abou 56 per cen of maize area is under hybrids, where maize yield is more han naional average. Therefore, assuming he before-menioned analysis as scenario-i viz. Business As Usual (BAU) scenario, wo more scenarios have been assumed for forecasing of maize producion a he naional level. The forecass presened under hree alernae scenarios, given in Table 3, indicae ha even if area under maize seizes o expand and effors are made o increase area under he hybrids, he oal maize producion can ouch 44 M by 2020 (Scenario III). This is only possible when overall hybrid maize yield improves o 5 /ha by he year 2020. However, if maize area expands along wih he adopion of hybrids a bi slowly (Scenario II), hen he oal maize producion may reach 37 M. This may offer a modes surplus of 5-10 M of maize for expor. However, i is perinen o menion ha wih expeced maize producion o be doubled in nex 7-8 years also require significan improvemen in logisics and supply chain like sorage, warehousing, ransporaion and grading and processing. Moreover, wih he recen implemenaion of Naional ood Securiy Ac (2013) and ongoing Naional ood Securiy Mission (NSM) may pu pressure on maize crop and overall area under he crop may no increase furher. Thus, here is a need o increase he adopion of hybrids and hereby, maize yield in much faser way han he exising pas rends. This inflecion poin will help he counry o mee he fuure growing demand of poulry and livesock feed as well as ap he opporuniy in expor marke. Currenly, poulry secor is growing by 9 per cen per annum and expeced o coninue in he fuure. Growing demand from poulry and feed secor in India is expeced o significanly hike maize consumpion o go over 30 M by 2020 [Raanani (2006), alcon (2008), ASSOCHAM (2009), Chennakrishnan and Raja (2012)]. The projecions show a surplus producion in he counry, which can be expored o he Souheas Asian counries. Similarly, here is huge maize expor marke of he size of abou 50 M in Asia iself, in which India is conribuing 4 M only in 2012 [UNCOMTRADE (2013)]. Wih improvemen in he yield and hereby reducing he cos of producion, he grain can be inernaionally compeiive.

Table 1 : CAGR (%) in APY of maize in major maize-growing saes of India. Sae Producion Performance of Maize in India : Approaching an Inflecion Poin 247 Area Producion Yield 1991-2001 2001-2011 2006-2011 1991-2001 2001-2011 2006-2011 1991-2001 2001-2011 2006-2011 Andhra Pradesh 5.63 6.32 2.78 10.11 11.64 8.93 4.25 5.01 5.98 Bihar -0.76 0.38-1.64 2.30 0.29-5.02 3.08-0.09-3.44 Chhaisgarh N A 0.95 0.32 N A 5.23 1.86 N A 4.24 1.53 Gujara 1.32 0.78 0.32 1.89-2.74 13.73 0.56-3.50 13.36 Himachal Pradesh -0.45-0.29-0.33 1.05-0.01-5.11 1.51 0.28-4.79 Karnaaka 10.53 9.90 7.89 10.19 13.79 10.04-0.31 3.54 2.00 Madhya Pradesh -0.43-0.79-2.11 2.10-6.02 2.86 2.54-5.27 5.08 Maharashra 8.74 11.49 8.68 7.45 16.22 14.70-1.19 4.25 5.53 Odisha -7.27 7.47 11.67-2.12 19.27 22.57 5.56 10.99 9.76 Punjab -1.81-1.88-4.85 1.56 2.88-1.58 3.43 4.85 3.44 Rajashan 0.23 0.96 2.57 2.97 3.43 7.08 2.73 2.44 4.39 Tamil Nadu 11.19 14.46 12.83 11.49 33.44 18.52 0.26 16.58 5.04 Uar Pradesh -1.67-1.77-3.80-0.03-1.17-1.65 1.67 0.61 2.23 Ohers 1.34 2.42 0.08 0.32 2.71-1.96-1.00 0.28-2.04 India 1.17 2.98 2.07 3.74 6.01 6.21 2.54 2.93 4.06 Table 2 : Disribuion of maize area and producion on he basis of growh and insabiliy in yield in India. Yield Insabiliy (CDI) CAGR in yield (2001-2011) Negaive Slow (0-3%) Medium (3-6%) High (6% & above) Low (<10%) Bihar (7.44%, 7.75%) Ohers (10.36%, 7.46%) Maharashra (8.99%, 9.71%), Odisha (1.00%, 0.98%) Punjab (1.66%, 2.50%) Medium (10-20%) Madhya Pradesh Uar Pradesh (9.12%, Andhra Pradesh (9.83%, 19.31%) (9.84%, 5.49%) 5.83%), Himachal Pradesh Karnaaka (14.5%, 18.22%) (3.54%, 3.26%) Chhaisgarh (1.21%, 0.73%) High (20-30%) Gujara (5.91%, 3.43%) Rajashan (13.12%, Tamil Nadu (3.49%, 8.64%) 6.70%) igures wihin parenheses indicae percenage of oal maize area and oal maize producion in India, respecively. - shows ha no sae falls under he respecive combinaion of growh and insabiliy. Source: Compued from he daa of Minisry of Agriculure, Governmen of India. Table 3. Projecions of APY of maize in India based on hree scenarios. Year Scenario I Scenario II Scenario III (Business As Usual) (Area growing as usual, adopion of hybrid (No area growh, adopion of hybrid reaches 75% & average yield from hybrid reaches 90% & overall average adoped area ouching 5 /ha in 2020) yield ouches 5 /ha in 2020) Area Producion Yield Area Producion Yield Area Producion Yield (M ha) (M) (/ha) (Mha) (M) (/ha) (Mha) (M) (/ha) 2012* 8.71 22.23 2.55 8.71 22.23 2.55 8.71 22.23 2.55 2013 8.77 23.34 2.66 8.77 23.72 2.70 8.71 23.86 2.74 2014 8.83 24.24 2.75 8.83 25.29 2.87 8.71 25.72 2.95 2015 8.86 25.09 2.83 8.86 26.96 3.04 8.71 27.83 3.20 2016 8.90 25.89 2.91 8.90 28.75 3.23 8.71 30.24 3.47 2017 8.92 26.63 2.99 8.92 30.68 3.44 8.71 32.98 3.79 2018 8.94 27.31 3.05 8.94 32.76 3.66 8.71 36.09 4.15 2019 8.96 27.91 3.12 8.96 35.00 3.91 8.71 39.62 4.55 2020 8.97 28.45 3.17 8.97 37.43 4.17 8.71 43.63 5.01 Source: Auhors esimaion. Noe: *Daa for 2012 are acual observaions for area, producion and yield as per 4h advance esimaes by Deparmen of Agriculure & Cooperaion, Gov. of India.

248 Ranji Kumar e al. 6. Conclusion The phenomenal growh in he producion and is spread across he regions proved maize a golden grain in India. Is diversified usage as food, feed and oher mulifarious indusrial derivaives make he crop special and apar from any oher cereals. Hihero, he increase in producion has been due o expansion in area under he crop as well as slow and seady improvemen in is yield. Maize in India has explored new regions and seasons, where i is performing relaively faser, while in radiional bel, i sared shrinking. Sagnan or negaive producion has been observed from Bihar, Madhya Pradesh and Uar Pradesh. Conrary o i, maize producion and produciviy has improved significanly in Andhra Pradesh, Karnaaka, Maharashra and Tamil Nadu. Though, 56 per cen of he maize growing disrics spread across he 25 Saes are harvesing less han 2 /ha yield. Besides, he curren naional average yield of 2.55 /ha is quie low as compared o ha in any oher Asian counries. This calls for sraegic change in he approach and give a posiive shock o he sysem so ha he resulan inflecion poin can add beer yield growh in he fuure. If he pas paern coninues, he maize area, producion and yield are expeced o be 9 Mha, 28.5 M and 3.2 /ha, respecively by 2020, while based on he alernae wo scenarios, i is expeced ha if he produciviy is improved o 4.2 /ha and 5 /ha, he oal maize grain producion may reach o 37 M (in Scenario-II) and 44 M (in Scenario-III), respecively. These forecass, however, would warran o have concered effors in developing and making available qualiy hybrid seeds suiing o differen agro-climaic regions. The oulook also pu bigger challenges o deck up he enire supply chain in maize secor o evacuae he almos double producion o avoid any glu in he marke. Moreover, wih he congenial policy environmen, India can paricipae in maize expor marke in a big way in he coming years. Acknowledgemens The paper is based on he findings of ongoing research projec on Assessmen of he maize siuaion, oulook and invesmen opporuniies in Asia funded by CIMMYT, Mexico. NAARM, Hyderabad is he lead cenre of he projec and oher counry parners include from Bangladesh, China, Indonesia and Pakisan. Therefore, he research eam duly acknowledge he financial suppor provided by CIMMYT, Mexico under MAIZE-CRP. References ASG (2012). Agriculural Saisics a a Glance -2012, Direcorae of Economics and Saisics, Deparmen of Agriculure and Cooperaion, Minisry of Agriculure, Gov. of India, New Delhi. ASSOCHAM (2009). Growh in poulry farming will push maize consumpion 30 M by 2020. Available on hp://www.dare.co.in/ news/ohers/growh-in-poulry-farming-push-maizeconsumpion-30-m-by-2020-assocham.hm (verified 17.07.2013). Barle, A. A. (1993). The arihmeic of growh : mehods of calculaion, populaion and environmen. Journal of Inerdisciplinary Sudies, Vol. 14, No. 4, March 1993. Boiroju, N. K. (2012). orecasing oreign Exchange Raes using Neural Neworks, Unpublished Ph.D. Thesis, Deparmen of Saisics, Osmania Universiy, Hyderabad. Box, G. E. P., G. M. Jenkins and G. C. Reinsel (1994). Time series analysis forecasing and conrol, 3rd ed., Englewood Cliffs, N.J. Prenice Hall. Chennakrishnan, P. and K. Raja (2012). Maize producion in India: fighing hunger and malnuriion, acs for You, November, 2012. Cuddy, J. D. A. and P. A. Della Valle (1978). Measuring he insabiliy of ime series daa, Oxford Bullein of Economics and Saisics, 40(1), 79 85. Dass, S., A. Manivannan, J. Kaul, A. Singode, J. C. Sekhar and G. K. Chikkappa (2010). Inbred-hybrid echnology in maize, Technical Bullein No. 2010/1, Direcorae of Maize Research, New Delhi. Available online a hp:// issuu.com/kisanadmin/docs/ inbrid_hybrid_ehcnology_in_maize_-mani (accessed on 10.08.2013). DMR (2012). Maize producion echnologies in India. Direcorae of Maize Research, New Delhi. alcon, W. P. (2008). The Asian maize economy in 2025. In: Gulai, A. and Dixon, J. (eds.), Maize in Asia: Changing Markes and Incenives. Academic oundaion, New Delhi, pp. 435-456. AOSTAT (2013). ood and Agriculure Organizaion Corporae Saisical Daabase. Available online a hp://faosa3.fao.org/ faosa-gaeway/go/o/browse/q/qc/e. Gulai, A. and J. Dixon (2008). Maize in Asia : Changing Markes and Incenive, Published by Academic oundaion, New Delhi. Haykin, S. S. (1999). Neural Neworks: A Comprehensive oundaion, Upper Saddle River, N.J., Prenice Hall. 16. Kasra, I. and M. Boyd (1996). Designing a neural nework for forecasing financial and economic ime series. Neurocompuing, 10, 215-236. Kaarkandi, B., N. K. Soora and P. K. Aggarwal (2010). Simulaing impacs, poenial adapaion and vulnerabiliy of maize o climae change in India. Miigaion and Adapaion Sraegies for Global Change, 15, 413-431. Raanani, J. (2006). India o see remendous changes. World Poulry, 22(6), 10-12. Tripahi, A. and A. R. Prasad (2009). Agriculural developmen in India since independence : A sudy on progress, performance, and deerminans. Journal of emerging knowledge on emerging markes, 1(1), Aricle 8, p. 63-92. UNCOMTRADE (2013). Unied Naions Commodiy Trade Saisics Daabase. Available online a hp: //comrade.un.org/ db/ (verified 18.07.2013). USDA (2013). India- Grain and eed Annual, 2013, USDA-oreign Agriculural Service, GAIN Repor No.IN3012. Zhang, G., B. E. Pauwo and M. Y. Hu (1998). orecasing wih arificial neural neworks: The sae of he ar. Inernaional Journal of orecasing, 14, 35-62.