65CJiO APARTADO AEREO 67-13 CABLES, CINATROP CAU - COLOMBIA I ~:.' ·\· ¡ -'U ' CENTRO INTERNACIONAL DE AGRICULTURA TROPICAL ¡l a Pulver Edware-oetic.~i-~e.e--'1'i.n,( (jf,ltice ',' - ',-' - ,-"- - ,- .", - '-'" r ! - - --- '-0;" ~.-' _ .• --0:- - '-- - ,,--' ------~- USE OF ANTHER CULTURE I Edward, PIl~ver 1/ "--~."'- ., l. INTRODUCTION • The production of doubled haploid rice plants through the culturing of anthers ls a well establ1shed laboratory technique that has numerous uses in rice breeding. The most immediate and perhaps significant application of this method i8 as a practical breeding tool since homozygosity can be obtained in less than one year after culturing anthers from an F1 planto The advantages of using the anther culture method for breeding purposes have been recognized for several years; however, a tremendous gap still exists between potential and practice. Some success as measured by the release and acceptance of new varieties has been obtained by several Chinese institutes working with japonica types with known androgenetic capabilities. However. in spite of these notable achievements, anther culture still remains to be incorporated into established rice improvement programs as a routine breeding method. • , • • 1/ Work partially financed by The Rockefeller Foundation. 11 Rice Program, CIAT, A. A. 6713, Cali, Colombia. / ....0 ,. \. \ ' ,.-.' A ' 2 Several factors have impeded the use of anther culture for breeding purposes. lt 1s well recognized that moat genotypea are not capable oí producing callus tissue from microapores with subaequent regeneration of doubled haploid plantlets. As a result, most research in anther culture • has been restricted to a narrow range of germplasm selected primarily for androgenetic properties rather than for their usefulness as parents in a breeding programo Genetic material that is of interest to the rice breeder has been largely neglected. Thus. anther culture is still viewed by most plant breeders as a laboratory curiosity. Another restricting factor ia the lack of suitable methoda to accommodate the large volume of material essential for a breeding programo Most rice breeders are unfamiliar with anther culture, although 1t is understood by tissue culturalists who are not sufficiently acqusinted w1th rice breeding to develop the required methoda. Thus, success in developing anther culture as a breeding method depends largely on overcoming these limitations. Anther culture research at CIAT (Centro Internacional de Agricultura Tropical) commenced after a considerable amount of information was available, thus enabling the work to be organized on an i applied baBis. The division between laboratory research snd field breeding was avoided by assigning responsibility for anther culture to • the rice improvement programo Work is conducted in collaboration with tissue culturalists and physical facilities are shared but the direction of anther culture research and day-to-day activities are the responsibilities of the rice improvement staff. This has resulted in sn 3 1ncreased awareness by the breeders of the usefulness of anther culture wh1le anther culture act1v1t1es are d1rected from a r1ce breed1ng perapect1ve. Research 1s restricted to genetic material employed in the breeding program resulting, initially, in low ratea of callus induction and plantlet regeneration. Work has concentrated on 1ndica rices although their ability to undergo androgenesis 18 much less than that of japonicas. However, much progress has been made and the doubled haploids produced using this approach have immediate commercial use. i 4 11. OBJECTIVES OF ANTHER CULTURE RESEARCH Tbe rice program at CIAT has regional responsibility for assisting national institutes in improving rice production in the Americas. Due to the tropical location of the breeding program, our ability to Serve • temperate areas is restricted. However, there are severe needs for increased rice production in the temperate areas of Brazil, Argentina, Uruguay, Paraguay and Chile and during the cold season in Cuba. Breeding programs in these areas are limited to one generation in the field per year and as a result progress is slow; requiring approximately 15 years to produce a variety. In most of the rice area in the Southern Cone, ta11 semi-improved varieties are still cultivated i.e. "Bluebelle" in southern Brazil, Uruguay, Paraguay and "Fortuna" in Argentina. In Chile, European varieties introduced more than 50 years ago are sti11 grown even through they are of poor quality and their market value is less than 50% of good quslity rice. Cuba is also a targeted area for anther culture. The country has a yearly production of approximately 500,000 tons of paddy rice but imports another 300,000 tons of milled rice annually. Varieties produced at CIAT are unadapted and cold-tolerant germplasm obtained from Asia has unacceptable grain quality. Attempts to combine the cold • tolerance of Asian japonicas with the indica grain quality have not been • auccessful. AIso, Cuba has a serious hoja blanca virus (HBV) problem. To incorporate these three characters plus blaat resistance into a single genotype would require 12-15 years using conventional breeding methods. 5 Anther culture breeding can be of tremendous vslue in temperate areas. The creation of homozygous lines in 7 to 8 montha would reduce the time required to produce a variety from 15 years to only 5 and the potential impact of a successful program would be substantial. In Chile, the identification of cold tolerant material with good grain properties would have a commercial value of at least US$10 miUion annually. In southern Brazil approximately 200,000 has can be incorporated into irrigated production if suitable cold tolerant varieties can be developed. This new production could eventually contribute up to 1 million additional tons of rice, essentially equal to the amount that Brazil 1s forecasted to import this year. The market value of this potential harvest would exceed US$300 million annually. Appropriate varieties for Cuba could almost eliminate the need for importatíon resulting in sn annual savings of approximately US $100 million. Breeding via snther culture could also be useful in our scid soil, upland program, essentially restricted to one growing season per year, by significantly sccelerating the development of improved material. Furthermore, the upland program is employing new sources of disease resistant material which eventually will be useful in the tropical irrigated programo The identification of new dwarf plant types from the upland program for use in the irrigated sector would broaden the genetic • base of our currently narrow irrigated germplasm which is exclusively of Asian origino During the past year, activities concentrated on anther culture 6 methods suitable for use in a breeding progrsm, identification of agronomically useful material that responds to anther culture, snd process1ng crosses that have resulted in the product1on of several thoussnd doubled haploid plants now being tested for their agronom1c usefulness in areas of potential use • • 7 111. STUDIES ON METHODS a. Plating of Anthers The most laborious step in the anther culture process is the • plating of anthers on the induction medium. The most commonly employed method consists of tapping cut florets on the perimeter of the induction flask thus permitting anthers to fall on the induction medium. With this technique one technician can plate approximately 1,500 anthers/day. This effic1ency 1s acceptable when the plant1ng of the Fl plants being processed can be staggered. However, the exposure of the Fl to selection pressures results in a volume of material that can exceed the capacity to plate anthers. At CIAT. all Fl triple crosses designed for tropical America are evsluated for HBV resistance. The evaluation is conducted only twice per year since rearing suff1cient vectors (Sogatodes oryzicola) of HBV 1s difficult. Consequently. we often have up to 400 Fl crosses available for anther culture processing during approximately 2-3 weeks. Current methods only allow for a amall percentage of the material to be processed by anther culture. Similarly, we observe that R2 lines derived from F2 plants previously evaluated for agronom1c characters and disease • resistance are much superior to doubled haploids regenerated from unscreened Fl material. However. the volume of F populations 2 • available far surpasses the ab1lity to plate the anthers. Two methods that could significantly reduce the effort required to extract the anthers were studied. One procedure 8 conaiata of placing cut floreta directly into the induction med1um and w1th agitation approximately 2-3 anthera can be aeparated from the glumes with eaae. Although thia procedure ia lesa time consuming than the standard practice, bacterial contamination 1a a serious problem. A similar method, originally reported by Japanese scientists, consista of placing the cut floreta directly into the liqu1d induction med1um and with mild ahaking during the induction process the call1 are separated from the glumes. However, contamination is high often exceeding 50% of the samples. With either of the two methods sur fa ce sterilization oi the glumes to reduce contamination has been only partially efiective with both methods. Our evidence indicates that the source oi the contaminant may be inside the florets. Current atudies 1nvolve the uae oi strong surface sterilants combined with selected antibiotics incorporated 1nto the induct10n medium. l'rel1minary results with this procedure 1ndicate that eontaminat10n can be redueed to an aeeeptable level but the effect of the aterilanta and ant1b10tics on androgenes1s on a range oi genotypes ia unknown. The ability to atore barvested panieles prior to plating would also faeilitate the processing of large volumes of material that is • available in a abort time. However, bacterial eontamination is also a serioua problem witb sto red samples, espeeially when the sp1kelets are left intact in the boot. We are eurrently examining various methods of treating penieles prior to atorage. Contamination is seldom mentioned as a eonstra1nt in snther 9 culture work; however. we have experienced serious problems. Some oi the contamination could be due to techniques, but s better explsnation may be the treatment of the Bource oi the anthers prior to collection. We extract anthers mostly from material after • disease exposure especially HBV and Piricularia oryzae. In contrast, most investigators report the use of greenhouse-grown plants, or, if field grown, in the absence of disease pressure. However, the benefits gained from culturing anthers only from screened material far outweigh the reduced contamination resulting irom growing unexposed plants. Consequently, our efforts are directed toward reducing contamination in preference to culturing material that comes from a clean environment. At present, we are limited to working with ireshly harvested material and gemplasm that has a rather high level oi callus induction ability. More research is required on methods of plating anthers to allow the processing oi more material, including germplasm that possesses only a low level of androgenesis. b. Induction Media The most stud1ed srea of rice anther culture 1s the influence of induction media on csllus formation. Unfortunately, mest of the work has be en conducted using a narrow japonica gemplasm base and • similar efiorts have not been devoted to indica types. The induction medium (including various homones) is important but it is not the controlling factor for callus formation. Induction media have quantitative iniluences on callus induction but the 10 media are not capable of converting a non-responsive genotype into a highly responsive one. .. Initial results at CIAT suggested a sign1ficant genotype X induction medium interacciono However. more detailed studies did not support these earlier observations. Results using anthers extracted from 24 hybrids of diverse crosses clearly demonstrated that the po tato extract medium containing 4 ppm of naphthaleneacetic acid (NAA) and 1 ppm kinetin waS superior to the same basal medium but w1th 2,4-D as the auxin or the N6 medium containing NAA and kinetin (Table 1). The interaction of genotype X medium was not significant. In 22 out of 24 hybrids the potato extract medium with NAA resulted in more calli than the other two media and in the remaining 2 hybrids it waa equally as good as the same basal medium with 2,4-D added (Fig. 1). Consequently, we employ Che potato excracC medium with 4 ppm NAA and 1 ppm kinettn throughout. New. media are being evaluated but it is not anticipated that a vastly superior induction medium will be found. c. Influence of Temperature Pre-treatment The use of a cold shock to stimulate androgenesis of japonica , types ia well documented; however, we question this method for tropical indicas having pollen sensitive to low temperatures. Anthers from three tropical indica varieties and one japonica were subjected to various temperature pre-treatments and incubated at 2So, 3D· and 3S·C. The results show that callus induction of the tropical variety "CICA 8" is increased by pre-treating the anthers 11 at S·C for 15 days (Table 2). "Fanny" (jsponica type) required only a 10 day treatment and 15 days of cold shock wss similar to the un-pretreated check. Pre-treating anthers at 30·C for only 5 • days drastically reduced androgenesis in all varieties. Sbort exposures at 3S·C or 40·C eitber bad a negative effect or no influence on the indica types. but signif1cantly reduced callus formstion of Fanny. Tbe incubation temperature ie al so important as temperatures above 2S·C completely inhibited callus induction in both the indicas and Fanny. The influence of low temperature pre-treatment on callus induction appeara to be similar to the effect of the induction medium; tbat 1s. it is quantitative. Tbe use of cold shock can stimulate induction in material that possesses androgenetic propertles but it will not significantly alter induction of a low calli-producing genotype. d. Influence of Regeneration Med1um The regenerative culture medium appears to have less lnfluence on plant regeneration than tbe carry-over effects of tbe induction medium. Tbe commonly used media for regeneration are standard media used for calli from rice and other various crops as well as vegetative tissues. The sol1dified Murasbige-Shoog (MS) medium containing 1 ppm NAA + 4 ppm kinetin is our standard and we have not encountered media tbat are superior. The regeneration environment. botb cbemical and pbysical. 12 plays an important role in regeneration. Genetic factors are obv1ously more critical but tremendous variation for plant regenerat10n ie frequently observed even from call1 derived from genetically pure material. The source of this var1ation 1s unknown, but normally 80% of the calli that are capable of undergoing plant regeneration will exhibit a green spot within 10 days after transfer to the regeneration medium. This suggests that the development of the callus at the time of transfer is critical. Calli exposed to the induction medium for long periods seldom regenera te plants. Consequently. early transfer of calli to the regeneration medium increases the probability of regeneration. However. very smaU call1 die if transferred too early. Methods that assure surv1val of tiny call1 would enable the transfer of minute tissue and thus increase plant regeneration. Commercial grade agar results in high mortality of small calli whereas purer agar, as well as cotton, dramatically increase callus survival and the pereent of plant regeneration (Table 3). Regeneration is sensitive to various types of contaminants (metals. toxins. etc) and eare should be taken in media preparation. At present we are using the MS medium with 1 ppm NAA + 4 ppm kinetin semi-sol1dified with Gel-rite. Cotton support is equally good but difficulties are encountered in removing regenerated plants. Commereia! agar is not a satisfactory support medium for regenerating plants from anther-derived callus tisBue. 13 e. Uniform~ty of Regenerated Plants Individual calli are products of single microspores and are easily maintained separately on liquid induction media. However, some researchers e. g. IRRI and the Shanghi Academy separate the clusters of plantlets originating from a single callus into individual plantlets, even through these plantlets should be genetically identical. Although this procedure increases the number of regenerated plants, the genetic diversity remains constant., The increased number of regenerated plants, many of which may be clones, complicates the evaluation of R2 lines, especially for multiple factors i. e. grain quality, iron toxicity, cold tolerance, and disease resistance. We observe that plantlets derived from an individual callus are phenotypically equal. Callus dissected in the early "green spot" stage to produce several green calli regenera te identical plants. Also, seeds harvested in bulk from unseparated plants originating from one callus do not produce significant amounts of segregating R2 lines. Less than 5% of the R2 lines demonstrate off-types which are discarded at harvesting and phenotypically identical plants bulked. Consequently, all regenerated plants discussed here are from individual calli and exclude plantlets from the same callus. 14 :•: .Uu. JI'. !. U•• , • 8 ,IP:·:! I I ...... 1::: :::: "l . 1••I ¡: : t ... :: :: U 1:. ., m .. • • I • • : • .. • .. .-< tI l .."... o :••: : .t :: l.t: : t.: lf1 •• ¡t I 'I..t :. .i :: I m fI •• u •• •••• •.1• .• ••. I. .... :.! .••. =o . e 11 m í li il! ¡ii lit i : i Jtii!i II i lil 1" ~ .... lf1 :;. :::: •:•: :, :1:: :::: '" 'O e :: ···:1 :l.: :: •• , ;:.* :. ::: 'e" i l.: ::: : :: !!: t:: !i!:1 e ro o :••:•: ·a·:1 • •:•:•: •:•:•: :•:• •, •a•:•: *' •!•:•: : ! ••:•: 1:.:1 tt.: ' tt: I •t•t•: •:•:•: ••••t•t•: l~ ....... o •••••••s•::• •• , •••••••••••••••••••••••• !I " '7 t:: :t: t: ::: s:: ::: ::: tt: :1: ::: tt: " lf1 '" N ttI: ::: 11' tI: ::: ::1 ,:: ::: 11: '1' :1: ,:: e 11 ___::_ __t_t : tt: s:: ::: t:: tt: tt: t:: tI: ::: :t: I "" ~---~----~~-------------------------------I Q) .Q..) "" e ~N~ ~N~ ~N" _N" _N" .N~ ~N" _N~ .N~ _N~ _N" _~" ....:.: lf1 N ~ 1 • .e... o.... .. .. .. N ¡ .. .. '.'"".. 11 i I .. : .:.;1 ~ .N. ~ . • ~ $ : .. ..... O :;1 ...... " v 111 '..". ': ... 11 N .•,1•.: .•: .•: ::• :: .:• 1.I ~ -,"., .a .t. i : : . i i i "N" Ijl I 1, i I i. ¡ II 1 1 · !l ..".. !••.•I 1• , :1 :• :I .!..'•• ¡t! .i :• e Q) .. ::,.:::::: '.:,: :' .. '• :¡' : :aaa. : I ..... lf1 !J . ::: l ...; o ....I. ¡EE lE . t¡ ,E ¡'Ii: f,¡t '¡' :1: I~ "" .... J••i ¡ ;:li l. ... . :' It:I •l .:I i. .1¡ .i .¡ . '.:•' ••1 o o ttI::: tt;I ' 1't1a tI 1"1':!: .1: ,.t::: ::1:: ::,: : m .. JI .. .... r'.". tI tI o .1::.:::. .1:.I:.t: ·:1·:1·:· :11::: .!t.:t.:! :. :::::: 1!t.1: :1:1:: • ¡.¡ :t:: ::1: m u Q) m e •••••••:•t•, .l. ,.. *. .• .•.• •••• •: •••• •: ...•. :.•: Q) OD ••••• •• 00 e :•t•· :. o .'; t•I: .•.•. : •.•.•• ••••• , : .•• ·.:t1t •I • : !:.,,:.. ,'·:t..· •s:•.• •:•:•::•:• ••:1• , ... ., .~ '" • :••:•: •t•·•· :•:• :·1· t• • ., . t: l' . ,.:.* , ,*: ¡: .J:": :•1• ,. :•:•:• •:•: : I 'm"c ! .J: ¡ " I ··t .: ••••• :. "O :1: ::t ~:t.:, .:.: : .t.I.: ..a. ::i :: ..: ...:1: tll .:1 t,:. .1 ::. 4-1 .'¡".l o ::: ::1 :1: :11 ~ .'".. -----------~---- .. ::: 1:. :1: ::: :11 tI: t:: :t: I .'; 11 _---~~---------------~---------- .'.". ... ... m '" 'Q" e ) ro • .. ~ . .. ~ I. ~ .. .. ¿ ~ ... ~ ~ : :.. Table l. Effect of induetlon medium on callus formation. Induetion Medium Rate of eallus Fomation l! Potato extraet + 4 ppm NAA + 1 ppm Kinetln 3.01 :!: .08 1:/ Potato extraet + 2 ppm 2.4-D + 1 ppm Kinetin 2.56 1. .ll N 6 + 4 ppm NAA + 1 ppm Kioet!n 1.79:!:.10 Seale 1 - S: 1 ~ 00 eallus; 2 = < 10; 3 ~ 10-25; 4 = 25-50; aod 5 = > 50 ealli/IOO antbera. y Values are meaos of 24 FI bybrlds witb a total of 162 observatioos followed by standard error • • 16 Table 2. Influence of temperature pre-treatment of anthers on callus formation of four rice varieties l! Variet;¡: 1/ Treatment Pre-treatment Incubation CICA 8 CICA 4 Or;¡:zica 1 Fann;¡: mean 1/ • Temp. ------------ No. of Calli/100 anthers ------------- None 25°C 8.0 3.3 0.6 338 87.5 8°C 5 days 25°C 10.8 0.7 0.4 415 106.7 8°C 10 days 25°C 12.2 0.3 2.0 504 129.6 8°C 15 days 25°C 27.5 O O 300 81.9 None 30 De O O O 11 2.6 30 DC 5 days 25°C O 0.8 O 1 0.5 30°C 10 days 25°C O O 0.4 4 1.1 30 DC 15 days 25°C O 0.2 O 1 0.3 None 35°C O O O 8 1.9 35°C 1 day 25°C 2.2 1.6 2.2 83 22.3 35°C 2 days 25°C 13.8 0.3 0.6 20 8.7 35 DC 4 days 25°C 1.0 6.0 0.4 62 17.3 40°C 6 hours 25°C 6.8 O O 131 34.5 40 Dc 12 hours 25°C 10.0 1.8 O 83 23.6 40°C 24 hours 25°C 2.6 0.2 O 98 25.2 1/ After pre-treatment anthers were incubated at 25°C for 50 days. 1/ LSD(.05) for comparisons of treatments within CICA-8 = 5.5; within CICA-4 1.4; within Oryzica-1 = 1.0; and within Fanny = 103. -3/ LSD(.OS) for comparisons of treatment means = 14.9 • • • Table 3. Effect of support media on callus survival and green plant regeneration. Callus Mortality Plant Regeneration --------------------------------------Support Media 1./ --------------------------------------------- Variety Ar COT Ag G-R Ar COT Ag G-R -------------------------------------------( % )--------------------------------------------------- TOX 1010-49-1 70 + 8.6 58 + 3.9 51 + 4.6 59 + 3.8 4.6 = 1.4 6.6 = 1.3 7.4 = 2.2 4.9 = 1.6 lAC 165 65 = 6.3 54 = 2.5 41 :!::. 4.3 55 :!::. 6.2 1.0 = 0.9 2.0 :!::. 1.3 3.0 :!::. 3.5 5.0 = 2.2 TOX 1011-4-1 63 + la 66 + 9.8 53 + 6.4 54 + 9.5 4.3 = 2.9 5.0 = 3.3 11.4 :!::. 3.4 5.7 =3.7 Colombia 1 68 + 4.8 38 + 7.0 63 + 5.6 48 + 12 o 12.5 :!::. 1.3 10.0 :!::. 5.1 13.3 :!::. 5.6 Col.lXM312A 97 + 3.3 40 :!::. 5.8 58 + 8.6 50 + 11 o o 2.5 + 2.0 o ..... Treatment " means 1/ 70 + 7.7 55 + 4.8 51:!::. 5.1 56 :!::. 6.1 3.3 + 1.3 5.8 = 1.5 7.2 + 2.5 5.5 :!::. 2.2 1/ Support media: Ar = Cornmercial grade Agar; COT = Cottoo; Ag = Agarosa; G-R = Gel-rite. Y Values are means of 64 repIications with 10 calli/rep. followed by standard error. 18 IV. Development of Cold Tolerant Germplasm a. Identification of Parents National programs in Cuba, Chile and southern Brazil as well , as CIAT have attempted for many years to combine the cold tolerance of some japonicas with indicas that possess good grain quality, but these efforts have been unsuccessful. Incompatibility between japonicas and indicas restricts conventionsl breeding efforts in this area. However, calli can be induced from pollen of crosseS that are highly sterile (Table 4), permitting the anther culture of crosses involving cold-tolerant japonicas with good quality indicas. The ability to induce callus in several known cold tolerant genotypes is shown in Table 5. The Chilean japonica varieties have excellent callus induction, often exceeding 900 calli/100 anthers. This is our most responsive material in terms of callus formation. Based upon their known cold tolerant properties and their androgenetic ability these genotypes Were selected as parents. b. Regeneration of Doubled Haploids The principal defect oí the Chilean germplasm is unacceptable • grain cooking quality. Consequently, the íirst set oí crosses was designed to combine the excellent grain quality oí - "Lemont" (US variety) that has modest cold tolerance with the highly cold tolerant Chilean germplasm. Three-way crosses were made combining two cold tolerant chilean sources with Lemont. 19 Material was processed via antber culture using tecbniques previous1y described. A summary of tbe sntber culture process is presented in Table 6. An arbitrary goal of approximately 150 doubled baploids/cross was set. However, sorne crosses were agronomically superior to otbers (1. e. CT 6741); consequently, more antbers were plated from tbese crosses. In general, tbe crosses were easily pro ces sed via antber culture due to tbeir bigh androgenetic ability. Callus induction as measured by tbe number of antbers plated was approxirnately 37%, and almost 9% of the call! were capable of undergoing plant regeneration. Further, approximately 3/4 of tbe green plants spontaneously doubled snd produced doub1ed baploids. Tbe same crosses were al so processed using conventional methods to compare tbe performance of lines derived from anther culture witb those from conventionsl breeding metbods. Heavy sterility "'as observed and under normal conditions tbese crosses would bsve been discarded. Only 234 F2 fsmilies were formed from 10 crosses. A comparison of the conventionsl breeding metbod witb antber culture i6 presented in Fig. 2. Higb selection pressure was made st CIAT for earliness, p1ant type and grsfn quality resu1ting in the identification of 190 R3 lines combining earliness, semi-dwarfism, long grain, clear endosperm, intermediate ge1atinization temperature and an amylose content of 21-25%. Similarly, approximate1y 270 F5 lines were derived using tbe conventionsl approacb. Due to tbe low number of 20 F2 families formed using the conventional method, the 190 R3 lines from anther culture may be more genetically diverse. The main difference between the methods 1s time; the 190 R3 lines were regenerated in 7 months whereas tbe F lines required four S generations. This material is presently being evaluated for cold tolerance in Chile and soutbern Brazil. Rowever, considering the parentage of the crosses we are confident that several of theae selections will exhibit excellent cold tolerance. Cold tolerant material with good grain and plant types is now available. To our knowledge, this is the first time that such material has been produced. This germplasm has tremendous immediate potential for Chile and otber temperate areas sucb as USA, and soutbern Europe and will serve as donors for cold tolerance in Cuba and southern Brazil for combination with RBV and bIaat resistance and iron toxicity tolerance. These new crosses are in progress. c. Soutbern Brazil The extreme south of Rio Grande do Sul requires varieties that possess cold and iron toxicity tolerance, bIast resistance, and earIiness. Approximately 100 triple crosses "ere made involving tbese combinations. Cbilean germplasm "as used as tbe cold tolerant donors and material from our tropical breeding program as sources for iron tolerance and blast resistance. Unfortunately, these wide crosses resuIted in aImost complete steriIity and less tban 20 crosses were processed using the conventional method. Anther 21 culture was not much superior as 60 of tbe 98 crosses faiIed to produce calli. Tbis 1s not surprising as irrigsted germplssm from tbe tropical breeding progrsm bas a very low level of androgenesis. Of the 38 crosses subjected to anther culture, 758 doubled baploids were regenerated but only 92 possess good grain quality combined witb earliness snd acceptable pIant type (Pig. 3). Tbese lines are presently being evaluated for iron tolerance and b1sst res1stsnce as well as adaptation to the conditions in southern Brazil. 22 Table 4. Relationship between callus induction ability and sterility of different F1 plants from four triple croases involving japonica and indica types. Cross: Diamante/IRGA 41011p2015-F 4 -66 Cross: Lemont/SI-21/Diamante Callus Callus Induction Sterility Induction Steril1ty ---------------- % -------------- -------------- % --------------- 29 98 30 98 26 98 27 98 9 98 94 95 9 98 58 95 138 95 20 95 41 95 66 90 14 95 100 85 1 95 47 85 39 15 68 60 28 15 84 50 Cross: Lemont/DiamanteIIP2015-F4-66 Cross: Lemont/IRI9743-25-21IDiamante Callus Callus Induction Sterility Induction Sterility --------------- % --------------- -------------- % ------------------ 13 98 31 98 6 98 2 98 10 95 32 95 100 80 2 95 O 80 14 90 78 70 7 80 1 70 6 70 80 40 3 50 16 30 O 50 3 30 O 15 • 23 TabIe 5. Callus formation ability of eold tolerant material seleeted from Asían nurseries and eold tolerant ChiIean germpIasm. Rate of Callus Entry Formation 1! I 32 1.3 SR 4079-4-2 1.0 Zho Fee No. la 1.0 China 988 1.0 China 1039 1.0 IR 9202-5-2-2-2 l. O K 31-163-3 1.0 K 39-96-1-1-1-2 l. O Tainung sin yu 1.0 Stejaree 45 5.0 Material from Chile Diamante 5.0 Q - 65101 5.0 Q - 64117 5.0 Q - 66304 5.0 Q - 67103 5.0 1/ Seale 1 - 5: 1 = no eallus; 2 = <: 10; 3 = 10-25; 4 = 25-50; and 5 = > 50 eaIIi/100 anthers. Table 6. Efficiency of 10 triple eros ses designed for Chile in the production of doubled haploids vi __ nther culture. Green Plant Effidencv in Rene~er;1tin~ P]ants Re~en~rated Production Douhlcd HilE,loinR Pedip,rE'e Cross N° of N° of % Total Creen Albinos í; % N' % !. % Anthers Calli lnduction Calli Anthers Doubled Green Calli Anthers Pl.ted Produced Hap10ids Planta CT 6741 Di_mante/LemontllQ 65,101 11000 4436 40.3 602 311 291 7.0 2.8 211 67.8 4.8 1.9 CT 6742 Q 64117/LemontllQ 65101 2900 1944 67.0 284 108 176 5.5 3.7 69 63.9 3.5 2.4 CT 6743 Q 65101/LemontllQ 65101 3000 491 16.3 94 53 41 10.7 1.7 28 52.8 5.7 0.9 CT 6744 Q 66J04/Lemontl/Q 65101 3700 2001 54.1 218 96 122 4.7 2.6 79 82.3 3.9 2.i CT 6745 Q 67103/LemontllQ 65101 1700 464 27.3 34 43 34 9.2 2.5 22 51.2 4.7 1.3 CT 6746 Dlamante/Lemont//Diamante 3100 1641 52.9 348 245 103 14.9 7.9 169 69.0 10.3 5.5 CT 6747 Q 64117/LemontllDiamante 2900 753 26.0 115 67 48 8.9 2.3 SO 74.6 6.6 1.7 CT 6748 Q 65101/LemontllDiamante 2500 1109 44.3 236 179 57 16.1 7.2 145 81.0 13.1 5.8 CT 6749 Q 66304/LemontllDiamante 7100 1046 14.7 184 123 60 11.7 1.7 116 94.3 11.1 1.6 CT 6750 Q 67103/LemontllDiamante 1200 496 41.3 87 56 31 11.3 4.7 52 92.9 10.5 4.3 Summary 39,100 14,381 36.8 2,202 1,281 963 8.9 3.3 941 73.5 6.5 2.4 25 lO Triple crosses 234 1281 F2 Families Ro Plants PLA NT/ TY PE I DOU8LE 98 SPS~ HAPrlDS F2 families ~ 941 QUALlTY Dihaploids 644 F3 Plants PLANT QUALlTY TYPE 207 SPS 1194 F3 Families F4 Seed QUAUTY 533 ./ R, Planto 670 F. . Unes I , PLANT TYPE PLANT TY!E 1 270 190 Ffi Unes R3 Seed CHILE Fig. 2. Flow of germplasm from crosses designed to combine cold tolerance with indica grain quality employing conventional breeding methods and anther culture. 26 98 Triple crosses ANTHER CULTURE NON-RESPONSIVE TO ANTHER CULTURE 38 Crosses I DOU8LE HAPflDS 758 Dihaploids 243 R, Seeds I aUALITY 92 ~~L.._:_:_¡_S_:_...J(~ Fig. 3. Diagram of eros ses programed for southern Brazil and proeessed using anther culture methodolOgy. 27 V. ADVANCES IN THE ACID SOIL UPLAND BREEDING PROGRAM a. Germplasm Evaluat10n The callus form1ng ab1lity of some commonly used parents from • the upland breed1ng program (ac1d s01l ecology) 1s generally greater than that of parents from the irrigated program (Table 7). Several genotypes from the upland program possess the ab111ty to form callus tissue which is not the case 1n the irrigated germplasm. Based upon this and other parental evaluat10ns, several crosses from the upland program were processed via anther culture. The object1ves of the upland program are to combine tolerance to soil ac1d1ty with 1ntermediate plant type and res1stance to HBV and other tropical d1seases. Resistance to HBV is controlled by a single dominant gene; consequently, the F1 of tr1ple crosses having one res1stant parent 1s evaluated for resistance. Th1s early evaluation allows for removal of the homozygous susceptibles, and anthers were taken only from the heterozygous plants after HBV screening. Theoret1cally, 50% of the doubled haploids regenerated using th1s procedure are res1stant to HBV. A frequency distribution for callus 1nduction and plant regeneration of 68 F1 triple crosses 1s shown in Fig. 4. More than 50% of the crosses exhibited sufficient callus format10n for processing via anther culture. Furthermore, about one-half of the crosses that possess androgenetic properties produced calli that resulted in greater than 10% plant regenerat1on. Crosses that exhibited acceptable levels of calIus formation and plant // - -:~--'. ~::-:~ .J .. 28 regeneration are listed in Table 8. Two commonly occurring parents in the better crosses are TOX 1871-38-1 and TOX 891-212-2-102-2-101 both of which possess h1gh levels of callus induct10n snd plant regeneration (Table 7). From these 68 crosses approx1mately 22,000 anthers were plated resulting 1n 8,000 cal11. While only 5% of the calli regenerated green plants, 69% of these were doubled haploids. Alb1nism was extremely high and appears to be common in crosses from the aeid soil programo The same procedure was repeated following the 1986 first season hoja blanca screening. Anthers were removed from Rr plants after virus screening in crosses that exhibited a high proportion of semi-dwarf plant types. The calIus forming ability of these eros ses 1s presented in Table 9. More than 50% of the crosses were capable of form1ng sufficient calIi to facilitate processing via anther culture and 15 crosses exhibited exceptional plantlet regenerat10n. Over 16,000 calli were obtained resulting in approximately 1,000 RI plants. We expect that about 600 these will be doubled hapl01ds. Anthers have also been collected from F 2 populations grown under extremely high blast pressure in the Colombian Llanos. This material was difficult to process due to high levels of bacterial contamination. However, 143 doubled haplolds were regenerated and most of these exhibit excellent plant type and good disease 29 resistance. Although a generation is lost by processing the F2 generation, the superior doubled haploids produced justify this procedure • • Advanced breeding lines from the upland program have also been evaluated for androgenesis. Of 16 F4 lines that possess semi-dwarf plant types combined with HBV and disease resistance four exhibit sufficient callus induction and regeneration to be considered as parents for use in anther culture (Table 10). These lines will be eros sed with selections from the cold-tolerant Chilean anther culture-derived material and doubled haploids will be regenerated for testing in Cuba. In summary, more than 1,000 doubled haploids have been regenerated from the upland breeding program (Table 11). These crosses were made independently of anther culture considerations, yet more than 35% of the anthers collected from a range of material produced calli. This clearly illustrates the high callus induction capability present in the upland germplasm which is largely African in origino However, only approximately 5% of the calli regenerate green plants. A comparison between the anther culture efficiency • of the Chilean and the upland material shows the higher rate of albinism in the acid soil crosses (Table 12). b. Perspectives The doubled haploids from the upland program represent germplasm that has tremendous potential. Many are semi-dwarfs 30 under irrigated conditions but maintain the deep-rooted character typical of upland material. Approximately 30% of the R lines 2 tested are also highly resistant to blast and 50% should be resistant to HBV. Furthermore, many of the lines are early, maturing in less than 110 days under tropical conditions and most are highly tolerant to iron toxicity. This new germplasm should be extremely useful in the irrigated program as it presents sources oí disease resistance and other characters (e.g. earliness) completely distinct from the currently available irrigated germplasm. Lines derived from the upland program will be crossed with the Chilean germplasm to combine cold tolerance with blast resistance and Fe toxicity tolerance to meet the conditions in southern Brazil. Also, the R2 lines with HBV and blast resistance will be combined with the Chilean material to fulfi11 the requirements for Cuba. These crosses will be processed using anther culture as all parents have known androgenetic abilities. 31 TabIe 7. CaIIus forming ability of anthers isolated from parents that are frequently used in the upland and irrigated breeding programs. • Upland Program Rate of CalIus Irrigated Program Rate of Callus Formation 1/ Forma tion 1/ COL 1 X M312A 2.1 IR 25840-64-1-3 1. O COL 1 2.1 GZ 864-2-3-1 4.0 -IRAT 122 1.9 ECIA 24-107-1 1. O IRAT 124 1.2 P 2867 F -52-2 4 1.0 TOX 1010-49-1 4.3 P 2887 F4-9-4 1.2 TOX 1011-4-1 2.3 P 3059 F4-25-3 1.6 TOX 1737-103-4 4.0 P 3299 F4-86 1.0 TOX 1768-1-2-1 1.8 16404 1.4 TOX 1768-1-2-2 2.0 17361 1.0 TOX 1780-2-1-1P-2 2.0 18521 1.0 TOX 1785-19-18 4.5 23916 1.0 TOX 1837-103-4 1.0 26444 1.0 TOX 1859-102-4M-4 1.3 11219 1.0 TOX 1871-38-1 4.5 l/ Scale 1 - 5: 1 = no caIlus; 2 = < 10; 3 = 10-25; 4 = 25-50; and 5 = > 50 calli/100 anthera. ". , .' " 30 25 25 20 20 c: c: .;o:; .;o:; ::l_ ::l _ :Q... *-- 15 :-Q... *-- 15 ti) 111 1/) e mm e .ro ~ 1nL r8888t E8888 ~ ~ L I em ~ mm 1m! m 51- ~ I§§§§I ~ ~ I§§§§I ~ 5 o O' ll888lI o >0-10 10-30 30-50 50-100 >100 <1 1-5 5-10 10-15 15-20 >20 (%) (%) Induction Regeneration of Green Plants Fig. 4. Frequency distribution for the ability to undergo callus induction and plant regeneration of 68 Fl hybrids from the acid soil, upland ecosystem. 33 Table 8. F1 hybrids from crosses made for the acid soil, upland ec08ystem that have acceptable levels of callus formation and plant regeneration. Pedigree Cross _________________________________________________ Induction Regeneration . CT 7488 IRAT 194-1-2-B/lRAT 122/ITOX 1780-2-3-2-1-2 30 14 CT 7422 NGOVIE/eOL 1 X M312A-74-2-8-8//TOX 891-212-2-102-2-101 25 11 eT 7415 TOX 939-107-2-101-1-1B/COL 1 X M312A//TOX 1780-Z-1-1P-4 23 12 eT 7413 TOX 1780-2-3-201-1/COL 1 X M312A//IAC 165 23 13 CT 7395 BR 51-282-8/eOL 1 X M312A//TOX 891-212-2-102-2-101 23 16 CT 7484 ITA 133/COL l//TOX 891-212-2-102-2-101 17 15 eT 7500 TOX 1010-49-1/IRAT 122//TOX 1871-38-1 16 15 eT 7411 TOX 906-2-1-202-2/COL 1 X M312A//TOX 1871-3B-1 15 17 eT 7469 ITA 133/COL 1 X M312A-74-Z-B-B//TOX 891-212-2-102-2-101 14 11 eT 7428 TOX 1010-22-7-16/COL 1 X M312A//TOX 891-212-2-102-2-101 12 19 34 Table 9. Callus induction ability of 48 Fl triple croases from the upland breeding program after screening for resistance to the hoja blanca virus. • Cross Rate of Callus Cross Rate of Callus Formation 1/ Formation ]J CT 7641 1.0 CT 7607 3.7 CT 7650 1.4 CT 7637 4.0 CT 8077 1.4 CT 8081 4.1 CT 7765 1.7 eT 7743 4.1 eT 8058 1.8 eT 7619 4.2 eT 8091 1.9 eT 7747 4.2 CT 7649 2.0 eT 7721 4.3 eT 7717 2.0 eT 7818 4.3 eT 7638 2.0 eT 7621 4.3 eT 7628 2.1 eT 7614 4.4 eT 8105 2.1 eT 7718 4.5 eT 7646 2.3 CT 8060 4.5 eT 7734 2.3 eT 7808 4.6 eT 8076 2.3 eT 7634 4.6 CT 7742 2.4 eT 8071 4.7 eT 7606 2.4 eT 7813 4.7 eT 7625 2.5 CT 7629 4.7 eT 8070 2.7 eT 7639 4.9 eT 7615 2.8 eT 7832 4.9 eT 7652 3.0 eT 8065 4.9 eT 7744 3.0 eT 7746 5.0 eT 8085 3.0 eT 7622 5.0 :. . eT 7656 3.1 eT 7633 5.0 eT 7642 3.2 eT 7833 5.0 lJ Scale 1 - 5: 1 ~ no callus; 2 = < 10; 3 = 10-25; 4 = 25-50; and 5 = > 50 calli/l00 anthers. 35 Table 10. Response to anther culture of advaneed dwarf, upland breeding lines that are resistant to hoja blanca virus and tropical fungal diseases, tolerant to iron toxicity and mechanical damage from Sogatodes oryzicola. • Pedigree Rate of Callus % of Calli Formation ~/ Regenerating Green Plants CT 6424-12-1-1 5.0 0.7 CT 6516-21-8-1 4.0 O CT 6516-21-8-2 4.0 9.0 CT 6516-21-8-3 3.0 O CT 6516-21-8-4 4.0 O CT 6516-21-8-5 4.0 5.3 CT 6516-21-8-7 4.6 3.3 CT 6516-23-10-4 3.0 3.3 CT 6516-23-10-5 5.0 8.6 CT 6516-23-10-12 3.4 2.5 CT 6516-24-6-2 3.5 O CT 6650-5-7-2 1.4 O CT 6650-5-7-3 1.5 20 Y CT 6650-5-7-7 2.0 O CT 6687-23-3-2 2.5 O CT 6687-23-3-3 3.0 O 1./ Seale 1 - 5: 1 no eallus; 2 = <10; 3 = 10-25; 4 = 25-50; and 5 = > 50 ealli/lOO anthers. ]j Data on only 15 calli • • 36 Table 11. Efficiency in regenerating doubled haploids from crosses for the acid 5011, upland ecosystem. Source No. of No. of % No. of Regenerated No. of • of Pollen Anthers Cslli Induction Green Plants Doubled Plated Produced Plants % Calli Haploids F - CIAT-85B 21,600 8,005 37.1 401 276 1 5.0 F - Savanna-85A 41,000 13,476 30.4 2 198 1.5 143 F3- CIAT-8SB 25,400 7,553 29.7 346 4.5 163 F - CIAT-86A 39.000 16.083 75.4 957 6.0 600 l! 1 Total 127 ,000 45,117 1,902 1,182 -_1/ Estimated: data incomplete • • 37 Table 12. Rate of albino and greeo plant regeneratíon and efficiency of doubled haploid production from cal11 derived from F¡ hybr1ds of crosses for the aeid soil, upland ecosystem as compared to Chilean crosses. Source of Re~enerated Plants Doubled Ha~loids anthers Albins Green Plsnts Total % No. % No. % Green Plants F1 - Upland-86B 3,019 76 945 24 582 61.6 F¡ - Upland-86A 13,330 93 957 7 600 62.7 F1 - Chile 963 42 1281 58 941 73.5 ". 38 VI. ACTIVITIES IN THE IRRIGATED PROGRAM a. Germplasm Evaluation The scarcity of genetic material tbat can produce callus ls the principal limltation for using anther culture in the irrigated • breeding programo Table 13 presents the callua formation properties for several high-yielding, irrigated varieties developed at IRRI and CIAT. None possessea sufficient callus induction abl1ity tor anther culture, nor do any oi the other Latin American (Table 14), irrigated varieties. Only "IAC 47", an upland variety, responds sufficiently to use it in crosses for processing via anther culture. These results are not surprising as nearly all Latin American material ls indica developed from a narrow Asian germplasm base. lf "IR 8", a widely used parent, possessed androgenetic capability then we would expect our commercial varieties to respond also. This is supported by data in Table 15. F5 lines trom atypical irrigated crosses (use oi HBV resistant donors trom japonica, African and upland sources) demonstrate some callus induction and three lines have a high ability to produce callus. The large difference between the responsiveness to anther culture pf the upland and the irrigated germplasm signifies that the upland germplasm is sUll genetically diverse or, conversely, the irrigated germplasm i5 extremely narrow. This serious limitation must be corrected to avoid disease epidemics. The newly produced upland dwarfs can significantly widen the genetic base of 39 the irrigated program and thus greatly reduce gene tic vulnerabili ty. 40 Table 13. Callus formation of sorne commercial irrigated varieties developed at IRRI and CIAT. 0_, • Rate of Callus IRRI Material Formation 1/ IR 5 1.0 IR 8 1.5 IR 36 1.0 IR 40 l. O IR 43 1.0 IR 50 1.0 IR 56 1.0 IR 58 1.0 IR 60 1.0 CIAT Material CICA 4 1.0 CICA 6 1.0 CICA 7 1.6 CICA 8 1.2 CICA 9 1.0 Metica 1 1.0 Oryzica 1 1.4 Oryzica 2 1.0 1/ Scale 1 - 5: 1 = no callus; 2 =< 10; 3 = 10-25; 4 = 25-50; and 5 = > 50 calli/lOO anthers. 41 Table 14. Response to anther culture of some eommercial varletles grown in Latin Amerlca. Variety Country Rate of Callus • where Cultivated Formation l! Anayansi Panama 1.0 Araure 2 Venezuela 1.0 Araure 3 Venezuela 1.0 Araure 4 Venezuela 1.0 Bamoa A 75 Mexleo 1.0 Bluebonnet 50 Several 3.0 BR IRGA 409 Brazil, Argentina l. O BR IRGA 410 Brazil, Argentina 1.2 BR lRGA 411 Brazil, Argentina 1.0 Campeche A 80 Mexico 2.5 Camponi Sur!nam 1.2 Ceysvoni Surinam 2.0 CR 201 Costa Rica 1.3 CR 1113 Costa Rica l. O Culiacán A 80 Mexico 1.0 Damaris Panama 1.0 Dlwani Surinam 1.0 Elon! Sur!nam 1.0 EMPASC 101 Brazil 1.8 EMPASC 102 Brazil 1.0 EMPASC 103 Brazil 1.0 EMPASC 104 Brazil 1.0 lAC 1278 Brazil 1.0 lAC 47 Brazil 4.7 lAC 165 Brazil 2.5 INIAP 7 Ecuador 1.0 INIAP 415 Ecuador 1.6 INTI Pero 1.0 Juma 58 Dominiean Republie 1.0 FA 2 Peru 1.0 Sinaloa A 80 Mexico 1.0 Tikal 2 Guatemala 1.0 • 1/ Seale 1 - 5: 1 = no calIus; 2 = < ID: 3 = 10-25: 4 = 25-50; and 5 = > 50 calli/100 antbers. 42 Table 15. Response to anther culture of irrigated selections possessing tolerance to Fe toxicity, resistance to HBV, tolerance to Sogatodes oryzicola. acceptable grain and good plant types. Pedigree Rate of CaUus Pedigree Rate of CaUus Fonnation -1/ Fonnation -1/ • P 5387-3-1-3-1 5.0 P 5746-18-11-1-5 2.0 P 5404-32-4-1-2 1.0 P 5747-12-9-1-2 1.0 P 5404-32-4-1-10 1.0 P 5747-12-9-2-7 1.0 P 5747-12-9-3-7 1.7 P 5413-8-3-1-1 1.0 P 5747-13-7-4-2 l. O P 5413-8-3-1-3 1.3 P 5747-13-7-4-5 1.5 P 5413-8-3-2-2 1.0 P 5747-13-7-4-7 1.0 P 5413-8-3-2-4 1.0 P 5747-13-8-2-2 l. O P 5747-13-8-3-1 1.0 P 5419-2-17-2-8 2.0 P 5747-13-8-3-3 1.0 P 5419-2-17-2-9 l. O P 5747-13-8-4-1 2.0 P 5419-2-17-6-1 1.0 P 5747-14-11-1-1 1.5 P 5419-2-17-6-3 1.0 P 5747-21-4-1-3 l. O P 5419-2-17-6-4 1.2 P 5747-24-5-1-3 1.5 P 5419-2-20-5-3 1.8 P 5747-24-5-1-4 2.0 P 5747-24-5-1-5 1.0 P 5690-1-4-1-1 1.0 P 5690-1-4-4-1 2.0 P 5690-1-4-4-2 1.5 P 5756-3-3-1-1 2.5 P 5690-1-4-4-3 1.0 P 5756-3-4-2-5 2.0 P 5690-1-4-4-4 1.0 P 5756-3-4-3-1 2.0 P 5690-1-4-4-5 1.5 P 5756-3-4-3-4 1.3 P 5690-1-4-4-6 1.0 P 5756-3-5-3-2 1.3 P 5690-1-11-1-7 1.0 P 5690-1-11-4-1 1.0 P 5446-6-6-1-13 3.5 P 5690-3-7-2-2 l. O P 5690-3-7-2-7 2.0 P 5583-4-12-3-1 4.0 1/ Scale 1 - 5: 1 ~ no caUus; 2 ~ < 10; 3 ~ 10-25; 4 ~ 25-50; and 5 ~ :> 50 calli/l00 anthers. • 43 VII. FUTURE ACTIVITIES a. Metbods for Processing Material A major constraint in using antber culture iB tbe inability to bandle lerge volumes of breeding material. Tbis i8 especially true • for crOSses tbat were not programmed for anrber culture, due to a bigb probability of encountering material witb low androgenesis. Furthermore, tbis problem i8 intensified wben material for antber culture i8 8imultaneously being evaluated for otber characters, e. g. HBV resistance. We are working on better techniques for plating anthers to reduce tbis constraint. It is more practical to pre-evaluate F 1 croases for callus formation and then submit only selected crosses to anther culture. Tbus, antber culture complements the conventional breeding program; tbat is, crosses tbat possess androgenesis are processed via anther culture and the non-responsive ones are bandled using conventional metbods. Using this method, efforts are concentrated only on androgenetic material to increase greatly tbe efficiency. The present procedure consists of planting 15 F¡ plants 50 days prior to planting tbe F 1 crosses in tbe HBV evaluation nursery and antbers are plated only from R resistant plants from crosses • possessing bigh levels of callus formation. • b. Development of Material for Cuba and tbe Southern Cone Germplasm sui table for Cuba must be cold-tolerant, possess tolerance to Fe toxicity, resistan ce to blast and HBV. These characters exist in two distinct germplasms: cold tolerance from 44 the Chilean R3 lines and the other characters trom the lines listed in Table 10. Donors that possess the desired agronomic traits and also respond to anther culture will be crossed for processing via anther culture. w A similar procedure will be used for southern Brazll. Cold tolerance will be obtained from the Chilean material and Fe toxicity and blast tolerance from the upland dwarfs recently derived via anther culture. Again, all parents not only fulfill the agronomic requirement but also possess androgenetic properties. These eros ses will be handled exclusively using anther culture. The material presently being evaluated in Chile has the qualities to produce new commercial varieties for that country following local evaluation and seed multiplication. No new crosses are programmed until information is available concerning the performance of this material. c. Irrigated Rice Program Tbe upland-dwarfs previously described are being incorporated into the irrigated crossing program to broaden the genetic base and to process selected crosses via anther culture. d. Development of National Anther Culture Programa Efforts have been made to strengthen facilities in Brazil. A small anther culture laboratory exists at the University of Pelotas, Rio Grande do Sul. Some of the croases programmed for the 45 Southern Cone ",i11 be processed in th1s laboratory. Ho",ever. infrastructure and personnel muat be expanded if anther culture 1s to be used for breeding purposes. Cuba also has expressed interest in establishing an anther culture fac1l1ty. A Cuban scientist has received tra1ning in rice anther culture at ClAT. 46 ACKNOWLEDGEMENTS Sincere appreciation i8 extended to Dr. Willy Roca, Head of Biotechnology Unit, as due to his efforts a rice culture program was established at CIAT. The assistance of Dr. John Miles and Dr. Roca in editing the manuscr1pt 1s acknowledged. The valuable techn1cal assistance provided by Mr. Victor Nuñez 1s apprec1ated. The author thanks Rosalba Reyes, Diego MacIas, Nancy Coral, Gerardo Delgado, José F. Ortega and Omar Herrera for the long hours of work involved in this project. Special consideration is given to Dr. César MartInez, Dr. Surapong Sarkarung, and Mr. Julio Holguln for assisting in the planting and selection of the breeding material. • • • - "