.~ ~~2.!!. 11/1/2.5 2 5 Ii,g = ::: 1/11/2.8 11111 . 1.0 3 2 1.0 3 2 W 1 . 2.2 I~ 1 . .2 I~ I~ ~ I~- ~ U~ w w I:Z E~ ~ e~ '".... ~ ... .. 1.1 .... &.;. ... 1.1 '"... a.- ....
-
111111.25 1111' 1.4 111111.6 111111.25 11111 1.4 111111.6
MICROCOPY RESOLUTION TEST CHART MICROCOPY RESOLUTION TEST CHART NATIONAL 8URtAU OF STANDAROS-1963-A NATIONAL 8UREAU OF STANOAROS-1963-A PRODUCTION OF
-: )") .) t , I•• ,-~ ~I l I ln J '-i I Technical Bulletin No. 1425 ... 1 ) I J.:.:. 't :5 I " -) 1.1 ) J.>-.{ .j -, : 1 U I~l
Agricultural Research Service . UNITED STATES DEPARTMENT OF AGRICULTURE ACKNOWLEDGMENTS The patient efforts of O. J. Eigsti, Chicago State Oollege, Ohicago, Ill., and Minoru Shiootsuma, Kiharn Institute for Biological Research, Yokahama, Japan, in the study and improvement of seedless watermelons deser'va generous recognition. Through personal visits find through correspondence 0 \Ter a period of several years, they have influenced the course of our l'e3earch and our conclusions. :Much credit also is due J. Robert
Wn.ll, whQ 1 while on the U.S. Vegetable Breeding Laborn tory staff, produced somo of rhe early tetraploid varieties and conducted ot,hm' research tha1tls basic to production of seedless watermelons. V. S. Seshadri is indebted to the Food and Agriculture OrglLnization of the United Nations for the Fellowship and to the US. Depn.rtment of Agri wltma (01' thc facilities that enabled him to participate in some of the studies. III CONTENTS Introduction__ -- ______Page1 Experimontal data_ -______2 Production of tetraploids______2 Production of 3X soo<1_ .. ______4 Flonting and gormination of 3X and 4X seed_ 7 Production of 3X and 4X Dlelons lmd their quality------'-- -______10 Discussion______11 Literature citod______12
'l'mde lULllles !Lre used in this publication solely for the purpose of providing specific information. iltfention of a trade name does not constitute [1, guarantee or warmnty of the product by the U.S. Depart ment of .A.griculture or an endorseme.nt by the Department over other products not mentioned. PRODUCTIOl'J OF SEEDLESS WATERMELONS
By O. .P. ANj)UUS, research horticulturist, U.S. Vegetable Breeding Laboratory, Plant Science Reseqrch Divi.~ion, Agricultural Research Service; V.S. SgSHAllRl, lrAO, jellow (India), Assistant P1"ojessor oj Iiorticlllture, Indian Agricultural Research Institlltc, New Delhi; P. O. GmMIlALL, research teehn-ician, U.S. lTeyelanle Breeding Laboratory, Plant Science Research Division, Agricultural Research Service
INTRODUCTION
Triploid (3X) hybrid watermelons, first produced by Killam and Nishiyttml1 in 1939 (3) / Me a c1tlSsicttl example of f;he use of colchicine induced polyplaids (OL' tho production of seedless fruit. In.panese scientists since then hn,ve tried continuously to perfect the technique n,ud make it n. commercin.l success (2,4,5,6). Serious problems remain so thn,t even in J tLpl1n, where In.bo!' n.nd cost of seed production are cheaper than in the United States, commercia} production of seedless hybrid wI1termeloll hilS not expanded as much as expected. Sevel'l11 11ttcmpts havc been made by certl1in State universities and seedsmen. to popularizc seedless wn.termelons ill the United Stl1tes, but they i1re still practicl111y unknown on most consumer mn.rkets. The U.S. V egetl1bleBreeding .Ln,bol'utory (VBL) at Oharleston, S.O., has worked on this problem since 1955 and three !LUtotetmploid breeding lines, Tetm 1, Tetm 2, and Tetra 3, were l'elensed in 1962 to seeclsrnen, who were to underLn,ke the production of triploid hybrid seed using suitable diploids (ZX) such as VBL 59-1 and VBL 59-6 relell,sed n,t the sume time. Seeclsmen, with few exceptions, fn.iled to respond to this opportunity, feeling thll,t too nULny problems of seed production and IUttrket production l'emttined unsolved to mttke the opern,tioll profitt1ble. Diploid hybrid watermelons lULVe been promoted by some commer cial seed producers at n. high cost to farm()rs. Triploid hybrid se()d are even more expensi\'e than orclinnry FI hybrid seeci, but the product hilS more potential I1dvt1ntages. Only a few seedless triploid hybrids hl1\-e been tested on the commercial m!Ll'ket; these came from Japn.nese seed producers Ilnd from the Americall Seedless VVn.termeloll Seed
I Italic ntlmbcrs in pnrenthcscs refer to Lilerature Oiled, p. 12. 1 2 TEC~ICAL n"VLLE'l'IN NO. 1125, U.S. DEPT. OF AGRICULTURE
Corporo,tion, Goshell, 1nc1. This U.S. firm has persisted with some success in efforts to salv:e the production problems with triploid hybrids, ILnd 11 few small centers of market production of seedless melons are beginning to appear. Because seedless and semise€:Jess watermelons offer all entirely different level of consumer appeal (1), additional eclucational efforts m'e needed i therefore, basing our conclusions on work n.t the VBL, we examine here in some detail the various problems involved in commer cil11 procluction of these specittl kinds of wl1termelon and suggest some prltcticnl solutions. 'VYe hop\} to interest extension wOlkers ilnd others in a position to promote overselLs markets lor fnrm products becl1use triploid fmel te1irltploicl (4)\) wlttennelons have specinl qlUt1ities Itdaptecl to this use. Among other n.dvfLlltnges, they luwe extrl1-hnrd rinds and extra firm flesh, which contribute to less dnmage in shipment and longer quality mainteUllnce for distant mn.rkets. Seedless fruit tend to keep longer, probably becnuse of reduced enzymlttic nction in the flesh at the IU'ell of unclmrelopeel seeel. The disaclvl1ntages, which include low producti\Tity nnd high cost of seed, 111'e no\\' partly solved l111d the way to better solutions seems near. Previously, personnel at the VBL have sought to demonstrate to reselll'ch workers Itnd growers the special qUl1lities of the tetro,ploid varieties of low seed content produced here, and some favorable re sponse is beginning to appear. Tetro, 2 melons are rated high in It few locations in India oml the Unitecl States where they have been tried. Seedsmen, ho\\'eYe1', nre reluctant to propagate a variety that has such a limited or Ul1l1Wllkened eleml1nd. "Ve here propose 11. method to COlll bine the yil'tues of tetmp\oids IUlCl the virtues of triploid hybrids as consumer products, ilnd inclicate how seed of both can be produced simultflIleollsly in the SIUIle field.
EXPERIMENTAL DATA Production of Tetraploids
The creation of 11 tetl'l1ploid variety is a prerequisite to the produc tion of seedless hybrids. While induction by colchicine in itself is compltrn.ti vely ensy lmel effecti \'e, the ll1ilintenance of tetrilploid water melon lines at 11 stable leyel is a required cflpltbility that must be proyecl experimenb1.11y. Strict vigilance is necessary to eliminate any clmnce oCC'UlTence of diploids thltt appltrently cnused the Jaj)nuese breeders SOlne difficulty. r1'he YBL tetraploids 11ltve been stnble in this respect and behnxe as true nu·ieLies. Clutnce diploid fruit CI1U be easily detected l1,t seed extraction by the greater number of seed, their PRODUCTION OF SEEDLESS WATERMELONS 3 smaller size, and distinctive pattern. A useful clue to recognizing a tetraploid fruit is the lIn'ger size of the blossom SCitr (areole) in relation to the smaller SCl1l' GL It diploid fruit. Ohance aueuploids cannot be distinguished by Itny Jpecific ciutrlLCteristic, but are extremely r!Lre. The seveml tetl'p,ploid lines e\Tolved at this Labomtory have been llutintl1ined 11t ltllLirly stable level by cllntinous mnss selection. They c0ntnill 11pproxinltttely 150 seed per fruit compared with 600 to 900 in the diploid. This level of seed fertility blS been built up through years of IJ11tSS selection frem eltdier counts of about 75 seed per fruit. Higher fertility in 11utotetraploids is possibly re1l1ted to lesser multivllJent fl'equoney in meiosis, which leads to rellsonn,hly true breeding behavior (6). 1i'or om purposc, scleetion for seed content (higher fertility) should not increase the seed numbel' to diploid levels but should stabilize it at fewer thn,n 200 sl;.cci per fruit. :Much lower seed content makes the cost too high from It seed producer's standpoint, Itnd much higher content defel1ts the purpose from a consumer's stll,ndpoint. The VBL lines hiwe close to the desired number of seed per fruit. Excellent [mit qtlltiity consisting of firm flesh, deep color, aud highly stable soluble solids content (12 Itnd 13 percent) is chltmcter isLie or the VBL tetrn;ploidlines. Attn,inment of superio!' fruit qtHLlity in f-lemiseedless tetrnploids is an important requisite for the production or high-qualit.y seedless triploids. II need for cau tion in the extraction of seed from tetmploid fruits is l'ocognized. li'ermentfttioll of the pulp should be ftvoided; the reduc tion in tetmploid seed vitality due to fermentt1tion seems well con firmed. High-pressure wlLter jets used with ftppropriate screens will effectively SeplLl't),te seed from fresh pulp, llnd most seed producers hl1\'e this equiprnent. In addition, we recommend that cOlnmercial producers of tetmploid (11nd triploid) watermelon seed subject the [I'eshly extrnctecl seed to n, compression process before drying. Tetra ploid seecl often dries with areas of ent1'l1pped 11ir and moisture, which It suitable pressing operation should help to reduce. The entrapp\~d air 11nci moisture in tetraploid deed is thought to have caused some of the eltrly
1il1O soed by Itir clt'ift. This procedure could also be ItdlLpted com t:Horeildly. vVe Imve l'egultwIy obtnined tetmploid seed of high gel'mi nnbility ILnd longevity sinco ndopting this technique. Seed villbility or ,-ito lity ILlso hns bnon indirectly improved through yom's of selection where 1I0nvillbie n,nd welLk seed were nn,tul'ILlly elimiun,teci. The size of tetl'iLploid seed is considered importlLnt by some breeders beClLuse the size of omplty ovules n,nd locules in 3X fruit is thoughti to be Im'Ge ill proportion to the size of seed of Lhe tetrn.ploicl femnle pltrent. This theol'}' hns not been fully coafh·med. Some smlLll-seedecl tott'nploids IlnN(\ been produced (l'om smiLH-seeded diploids ILnd otihers luwo beon solected frnm olttcrosses, but it romllins true thlLt some l!Lrge-seeded totrlLpioids WIl'minltte betLer u.nd produce strongor seedlillgs. 1Voj lu'e (\lllTontly recommending Tetl'l1, 1, Tetm 2, und Totl'l1 3 from the YBL, both ns pnrenLs in the production of seedless hybrids tLnd ns s!'lLble varieties tihomsel ves with high consumer n.tliributes (fig. 1). Tetl'lL J and 'l'etrll, 2 IL1'O indistinguishfLble I1S vllrieties but perforlll ciiffeL'ontily 11$ pn,ront.s. From a seed prOclllCel"S stancipoint t,hese all Illwc sufIkiont prociucLivil,y (2 fruit~] pOI' pli1llt) I1nd suffic\ont soed content. (150 pOl' melon) Lo mn,ke soed production possibl'd without excessi \'e eost. Newor V.BL LeLmploicis, ;,llCh IlS "V872 ILlld Vl,Tl068A, promise even botitor porfonnn.nce. Still other usoful Americl111 tetrt1 ploid sLocks 111'0 Iwn,ilable from other breoders.
Production of 3X Seed
Pl'Oductiion of triploid hybrid seed by the originallLnel costly process of hfllld pollitllltion of 4X x 2X parents must be reje(~ted for pructicn.l 1'onS0118, N n,fiuml erossing by boe POUilll1tioll in field plots seems the emeien!. 1'L1l8wor to seed producLioll, iLnd vl1rious field designs of mixed pln,ntillgs of teLrl1ploids I1nd diploids IHLVe boon tried. ShimotsumlL (6), in JILP11l1, goL best results with threo totrO-ploid rows to ono diploid row, ,,'het'ons "Vall U.,'), from the V.BL, reported 83.6 percent triploicls from 11l1turfll crOS8-pollinn,t.ion with one totl'l1ploid row to one diploid. The In,Llor 1: 1 design is in usc Ilt the V.BL n,nd seems to be suitn.ble with 6-foot. beds. If It 3: 1 design is used, i1 grel'Lter proportion of 4X seeds will be produced by self-pollinI1Lion. To insure higher proportion of 3X seeds, n.lt:cl'l1n.t:e (1: 1) planting of 2X l111d 4..,."'C is desimble (fig. 2, A). Outside (borcicI') rows of diploid parent. pln,nts cn,n be used for seed pI'odu('tloll but inside rows of diploid pnTent should be pulled im mediately llfler pollinn,[.ioll. This gi\'es tctmploid plttnts nlot'e sr;,\ee t.o sprcnd l1nd pre,-cu[:,; ('()n[u~i()n of Lht' l.\\-o pi1ronLI11 types of fruit, tLt lHLlTest. PHODUC'l'lON OF SEEDLESS WATERMELONS 5
DN-37215, DN-37216 FlGmn: l.-Tetrn 2 (top; and Tetra 3 (bottom), tetrnploid (4X) watermelon variot,ies relensed to commercial breeders by the F.8. Department of Agri cultlln' for \lilt' in chI' production of secdlesH hybrids, arc themilelves good consumer productH. Tetra 1 (not shown), Similar externally to Tetra 2, is nn equally good product. 6 ~'ECHNICAL BULJ~ETIN NO. 1'125, U.S. DEPT. OF AGRICULTURE
A. Generation for production of seed (2X, 3X. and 4X):
Plants Rows 1 Seed treatment 2
2X ////////////////////// Save seed to propagate 2X parent. 4X XXXXXXXXXXXXXX Save all seed from 4X rows. 2X ////////////////////// Remove these 2X plants after pollination. 4X XXXXXXXXXXXXXX Save a\l seed from 4X rows. 2X ////////////////////// Remove these 2X plants after pollination. 4X XXXXXXX~XXXXXXX Save all seed from 4X rows. 2X /111////////////////// Save seed to propagate 2X parent.
B. Generation for production of mEJlons for market (seedless and semi seedless):
Mixed 3X and (X)(X)(X)(X)(X)(X)(X) Leave all fruit for single harvest. 4X plants (X)(X)(X)(X)(X)(X)(X) 3X plants produce medium-green (X)(X)(X)(X)(X)(X)(X) melons. which are seedless. 4X (X)(X)(X)(X)(X)(X)(X) plants produce light·green melons, (X)(X)(X)(X)(X)(X)(X) which have few seed. 70 to 90% (X)(X)(X)(X)(X)(X)(X) of the melons will be 3X; 10 to 30% of the melons will be 4X.
1/1/ Indicate 59-1. dark·green watermelons; XXX Indicate Tetra 3. light·green watermelons; (X)(X)(X) indicate mixed 3X and 4X. These symbols have no sig· nificance as to number or spacing. They suggest an indefinite number of plants arranged In paraliel rows. 2 Seed from the 4X rows will be mixed 3X and 4)( and can be sold or used to prop· agate the 4X parent. the 3X seed will not reproduce itself.
FlOUln] 2.-Productioll of seedless wntermelons: A, Generntion for production of seed; B, generation for production of melons for mnrket.
Thus fo.r, seed lmrvested from 4X plo.nts (11rro.nged us in fig. 2, A) luwe proved to be 11pproximl1tely 33 percent 4X (sib-selfs) o.nd 67 pereent 3X (outcrosses). Attempts to separo.te the two kinds of seed beCore pla,ntings ho.ve had only slight success. Kondo (4) and Shi lllotsllmo. Imel .Mo.tsumoto (7) predented 0. sorting method of diploids, triplnids, and tet1'l1ploids on the busis of difference in tliickness and weighL of seeds, n,nd O. J. Eigsti (letters on fUe) found differences in specific gro.vity of the tihree kinds of seed. But in commercil11 produc ti.on, these differences do not lend themselves for prodico.l use; seed chnmcters have to be distinct enough for efficient sepal'l1tion in me cho.nical ope1'l1tiol1s. We recommend instead thllt the 3X and 4X seed be kept and sold ns fL mixture (fig. 2, B). This ussumes not only that both the 3X fl,nd the 4X melons to be hn,l'vested will be marketable but also that they call be sopt1l'fLted ltttor on the busis of external characteristics. PRODUCTION OF SEEDLESS WATERMELONS 7 In the ubsence of color lllurkings, the 3X fruit cun be distinguished from 4X with IL high degree of uccurucy by reference to its more triungulur (lobecl) shape when viewed from the blossom end. With some pructice, this shape can be 11 highly reliable indicl1tor. In recommending thu,t 3X l1ud 4X seed be retuined in their originul mixture of 2:1 (fig. 2, B), we ussume ruso thut the 4X plunts will be effective pollino,tors to cuuse fruit to set 011 the 3X plants. So far this nssnmption hus be6ll true in our experiments, provided there is no scurcit,y of bees. Eliminution of the 2X pollinator in the lllurket pro ductioll stuge is n, key fen,ture of this procedure because it greatly simplifies the routine for the grower.
Planting and Germination of 3X and 4X Seed The erratic germination of 3X ILnd some 4X seed is well known, and thus they should be germinated in fluts or pen,t cups lLud then trans plrmtecl in the field. Germinl1tion data from two experiments ure shown in tn,bIe 1. .Although trlLl1spln,ntil1g will incrense cost of pro duction of seedless fruits, it ruso provides llutximum economy of expen-
TABLE I.-Germination of 3X and J,.X watermelon seecl har-veste(l i'n 1967 ancl tested in September 1969
Experimental lots Seed Germination planted
WIOS3: Nltmber Nltlllber Percent 3X (W987) ______300 247 83 4X (Tctm :1) ______300 249 83 :i\-[ixed3X (hand______pollinated): _ 4X______150} 197 66 150 Mixed (bec pollinated): !~}------300 277 92 WI084: 3X C\V985) ______300 175 58 4X (Tetra 2) ______300 172 57 Mixed3X (hand______pollinated): _ 4X______I50} 167 56 150 Mixed (bee pollinated) : 3X1. 4XJ ------• -- 300 206 69 8 TECHNICAL BULLE1'lN NO. 1 1125, U.S. DEP1'. OF AGRICULTURE siye seed !l,nd is a practice nlreaely widely used by market gardeners for other crops. An n,ttompt wns made to compare direct seeding l1nci lil'l1l1splal1ting with respect to stl1nd Mel 111so to the proportion of 3X n,nd 4X plants I1nd fruit Jlroduced from the mixture. The popull1tion used WfiS bee pollinn,ted. The stand count datIL were too meager to report, but the lllLryest cin;ta m'e substn,lltinl I1nd ngreeltbly surprising (table 2). The proportion of triploid pln,uts was not, 11laterinlly different in the two methods, but sl;n,nd wns beLter I1nd the number of triploid fruits was greltlel'in trn,nsplal1 ted plots. "VIe could notdetel'mine whether the selec tion preS8ltl'e wns in fl1\'or of 3X or of 4..,."{ pll1nts when ungerminl1ted and excessively weak seedlings were elimiultted in tl'l1l1spln,nting. The agreeably surprising ctn.ta t.hat hlwe come from these tests is the high propOl'tioll of slll'vi vn.l of Lhe 3X plants from Lhe mixt;ure of 3X and 4X seed n,11d I,he high proportion of seedless (3X) fl'Lljt in the final. hal'vest;. '1'he 1l1LLer figure hilS someLiLnes exceeded 90 percent; (tl1ble 2 I1nd figure 3). PrcsUl,l1n,bly, this predominl1nce of 3X rell1tes to a remnant of l1utosterility in the tet1'l1ploids. Becallse trl1nsplanting seemed to be more efficient thltn direct field seeding, efforts were directed to find H,ny seedling character that, could be identified to sepl11'11te 4X seedlings; for even though 4X plants are to be retained ns pollinl1tol'S, it might be helpful to sepa1'l1te l;ilC 3X and 4X in alternate rows for ensier harvest separn,tion. Initial obsern1tions indicate t.hat pln,n ts with folded cotyledonary lel1ves may be up to 95 percent, Ll'iploids (table 3), while the converse is not true; all plnn ts \\'ith normal flltt cotyledons i1ro not necessllrily tet1'l1ploids (ttlbles 3 11nd 4). Hence, this seedling cha1'llcter, like the seed chal'!lctoJ'8 mentioned earlier, must be rejected as not sufficiently reliable.
TABLE 2.-F'ropartion of SX and ;'X seedless ~vatermelons harvested . (~fler bee pollination
l\lelons harvested VBL No. and method of seeding ~------3X 'lX Total
Number Percent Number Percent Number WI083: Direot seeded. _____ • 94 79 2'1 21 118 Tmllsplaute(L .._•. __ • _ " _.• 183 90 20 10 203 \V1mN,: Direct seeded _____ • "" ____ . _ . _ GO Se. S 12 68 TmllsplltlttccL ___ .."_ _. _. _ . . 9.1 92 7 17 102
..-~---~--~~,-.--~-- --- 1 2X=,\ melon: 1 percent. , .Iii. o~ ti'IJO.--.~.p;~~ ~ .."'Ili:"'Ili:7" tl c:: (') o~ Z o ";i
YJ t;j gt;j t;j YJ YJ
~ t;j ~
~ l:"
~ YJ
BN'-37217 FIGURB 3.-A, Cut fruit of seedless (3X) and semiseedless (4X) watermelon component from bee-pollinated population WI083; B, seedless and semiseedless melons of WI083 in the proportion (7:1) in which they occurred at harvest; C, cut fruit of seedless (3X) and semiseedless (4X) component from bee-pollinated population WI084; D, seedless and semiseedless melons of WI084 in the proportion (10:1) in which they occurred at harvest. The 3X and 4X melons in WI083 are distinguishable by color, but those ill WI084 are distinguishable only by shape viewed from the blossom end. to 10 TECHNICAL BULLJjlTIN NO. 1425, U.S. DEPT. OF AGRICULTURE rl'ABLE 3.-Proportion of 3X a'ncl ;"X melons derived from plants wit1b normal vers~ts folded cotyledons
Melons harvested Cotyledon shape Hills 3X 4X Total
NI!wber Number Percent NlImber Percent NlImber NormaL ______Folded______50 56 72 22 96 78 50 71 97 2 3 73
TABLE 4.-Folded cotyledons in relation to ploid1l1:n watermelons
Cotyledons VBL No. Ploidy Total Normal Doubtful Folded seed flat class lings
NlImber NlImber NlImber Percent Number W915A______2X______235 1 0 0.1 236 W987 ______3X______81 6 24 21. 6 111 Wl030______3X______63 7 20 22. 2 90 WI060______3X______68 12 98 55.1 178 Tetm 2 ______4X______322 5 7 2.1 334 Tetra 3 ______4X______324 3 2 0.6 329 W872-R______4X______224 1 4 1. 8 229 W972-0______4X______177 20 12 5. 7 209 WI068A-9-ML_ 4X______204 , 17 19 7. 9 240 WI083______Mix 3X and 198 8 31 13.1 237 4X. WI084______do ______171 9 59 24. 6 239 Wl111______do______155 9 65 28.4 229
PI'oduction of 3X and 4X Melons and Their Quality The 3X fruits do not develop parthenocarpically but must be stimulated by pollen. Apparently 4X pollen is just as effective as 2X pollen, but is less abundant; hence ample bees should be present. Provision for a diploid pollinatoi may be necessary for pollinating 3X flowers if bees are scarce and if there are not enough tetmploids in the mixture to meet the pollination requirements. Distinction of 4X and 3X fruit (fig. 3, A, 0) can be made in the field at hal'\Test with the help of rind characters such as color in Wl083 (fig. 3, B). Even when 4X and 3X fnlit are similar externally (ns in fig. 3, D), upon close examination, triangular conformation of 3X fruit is distinguishable from the more spherical conformation of 4X fruit. The quality of 3X fruit in the VBL materials has been good; tIfe PRODUCTION OF SEEDLESS WATERMELONS 11 bud features sometimes found in 3X fruit, such as hollow heart, hard seed coats, lllrge empty seed caviti"s, Ilnd brown seeds, have been found infrequently in the VEL mat6rials. The melons have good eating quality) which CIUl be appreciated by any discriminating consumer. Sweetness is notably higher and more stable than in the diploid parents. Sll.~dsmell can choose any particular combination, depending upon whether striped or unstriped fruit, round or oblong shape, or Inrge, rnedium, or small size is desired by consumers in any region. The smltllest VBL triploid (fig. 3) D) is in the Sugar Baby size class, rang ing from 9 to 17 pounds and averaging 11.7. At present (1970) a num ber of tetraploid lines luwe different rind characteristics and a suitable one, fulfilling the consumer preferences for certain external characters, Cltll be chosen. Additional diploid parent pollinators need to be tested to provide distinguishing characters in the 3X fruit. DISCUSSION Insufficient diseuse resistance hus been one of the limitations pre venting the wide use of triploids. Tetraploid lines with wilt resistance incorporated llre now produced at VBL and at Purdue University; also, many diploid parents like the VBL Wilt Resistant Oongo carry wilt resistance. Hence, good wilt-resistant tetraploid and diploid lines are now lwailable. Some of the newer breeding stocks like VBL W1068A !lIso carry some measure of anthracnose resistance under field conditions. The polyploicls seem even to have nn extra degree of field tolerance to disease when compared with common varieties. Hence, luck of disease resistance need not be n deterrent for commercial production of seedless watermelons. It is evident that some problems of production of seedless water melons have been surmounted, but some still remain to be solved. What is needed tocit"ty is un honest effort in the application of available findings. 'l'hough better solutions will be found, we need not wait for them before we make seeclless watermelons available to the consum ers. Oonsumers luwe long been accustomed to eating a particular kind of watermelon, like Oharleston Gray of the East or Klondike of the West, and the time has now come to put a new kind on the market. Buyers would easily perceive and accept the excellent quality of prop erly produced seedless melons and would not grudge paying a few cents more per fruit. Shipping agencies nre not bothered by what is inside the fruit; they are concerned only with external feattlres. The seedless and semiseed less :melons would ship find keep well because they have extra-tough rinds and only a few seeds to provido loci for overripening. In fact, the Oharleston triploicls and tetraploids are so stable in internal flesh 11
12 TECHNICAL BULLE'rIN NO. H25, U.S. DEPT. OF AGRICULTURE ! quality thn,t workers at the VBL leave these melons in the field for a single hltrVest. Henco, for large-scale production and long-distance shipping, seed less and semiseedless melons will not suffer on any account by compari son with similar products. The cost of production will be higher and, like any new variety or novelty, these melons will tend to cost more initially. But unless [t fair trial is given and repeated in many localities, growors will not bo Itblo to discern tho smail difficulties and improve tho production techniques. According to n,!. Shimotsuma (personal com munication), growers in Taiwan prod HCO seodless watermelons on a In,rge scale n.nd ship thorn to markets in Hong Kong and Singapore, growors in Spnoill ship othor kinds of molons to the United States, and growors in tho Middlo Enst ship melons to western Europe. A better Itppreciation of production techniques and a change in the psychology of soedsmen, growors, shippers, and consumers could definitely briu!:; 11 n LITERATURE CITED (1) ANDRUS, C. F. 10G9. 'WI> CONSUMI-:lt DgMAND Fon WATlmMBLONS. Veg. Growers ,Mes senger (April), p. 7. (2) Km,uu, II. 11)51. TlUl'f,QID WA'rlm~mf,QNS. Arner. Soc. Hort. Sci.l' roc. 58: 217-230. (3) --- and NISIIIYA~IA, 1. 19~17. AN Al'l'r,ICATION OF STBlUl,l'l'Y Or' AUTOTHIPLOID TO TlIEllHBEDIN(} O~' SI~I.;nfJgSS WATBHM"LONS. Seiken Ziho 3(III) : 5-15. (-1) KONDO, N. 1955. STUDms ON ~'Im TIUPJ,OID WATlm~mI,ON. lnst. Breeding Res. Tokyo t\gr. Unh'. 52 pp. (5) SllIlIIOTSUMA, l\I. ~')Gl. A SUHVI~Y O~' SI':lml,ESS WA'n~H:.n1LON 1IH\,I,D1NG IN JAI'AN. Sciken Ziho-Collectivc ,Revicw Report 12, Kiham Inst. BioI. Res. pp.75-84. (G) 1962. STUDlI~S IN TRIPI,om SI'lI~D I'ltODUGTION IN WA.T~lR:'U:LONS. Jap . •Jour. Breeding 12: 124-129. (7) --- and :\lATsuMoTo, K. 1957. COMPAHATIVI'; sTUDIgS ON ~'IH; ~IOUPIIQl,OGY OF POLYPLOID WATgH "!ELON SI':I;;o8. (,English summary] Sciken Ziho 8: G7-74. (8) \'hr,L,.r. IL ,WGO. mn: O~· MAllIU:tt G~:NI': IN {,HODUCING 'rHIl'fJOID WATJmMI~LONS. Amer. Soc. Hart. Sei. Proe. 76: 577-581. I! , , " .'