After reading this article you will learn about:- 1. Origin and Production of Cassava 2. General Description of Cassava 3. Breeding Objectives 4. Hybridization 5. Breeding Methods 6. Use of Double Haploids 7. Germplasm Conservation.
1. Origin and Production of Cassava:
Cassava has been studied since as early as 1886 when Alphonse de Candolle placed its geographic origin in the lowland tropical Americas. It shares the Brazilian-Paraguayan centre of origin with peanuts, cacao, rubber and other crops.
Cassava (Manihot esculenta Crantz 2n = 2x = 36) is considered to be one of the most important sources of energy for poor people living in the rural areas where cassava is grown and consumed as a staple food. On a dry matter basis, the calorie content of cassava/g of edible material is similar to that of rice and maize.
However, the total protein content of cassava is low while its quality is poor due to the low content of several essential amino acids, such as lysine, methionine and tryptophane. Despite these limitations, cassava still dominates over other starchy crops since the risk of crop loss due to diseases, pests and climatic uncertainties is lower than for other competing crops.
The world acreage of cassava is 16.37 million hectares with a production of 164.75 million tonnes. Nigeria occupies the first position in area accounting for 16.5% of the world area producing 18.5% of world cassava. Congo, Brazil, Thailand and Indonesia are the major cassava growing countries of the world.
In Asia, cassava utilization patterns vary from country to country. In Thailand, cassava is used for the production of starch and pellets and is not directly used as staple food, although small quantities are consumed in snack form.
In Philippines, it is primarily used for starch production and as a domestic food. In Malaysia, cassava is mainly processed into starch and some is used as a traditional food. Other countries growing cassava on significant scale arc Indonesia, China, Vietnam and India in Asia.
Among the root and tuber crops grown in India, cassava comes after potato in production and area. In India, the total cassava production is about 6.0 million tonnes from an area of about 0.24 million ha with an average yield of 25 t/ha, which is highest in the world, the average cassava yield in the world is about 10 t/ha.
It is grown in the southern states of India, viz., Kerala, Tamil Nadu and Karnataka and Andhra Pradesh. It is basically used as staple human food and its industrial use as animal feed and for starch production is gaining importance. Types of traditional food from cassava in use in different countries are shown in Fig. 35.1.
2. General Description of Cassava:
Cassava is a lactiferous shrub with upright stems marked by leaf scars, reaching to a height of about 5 m. The leaves are palmate and rarely seen flowers, resemble a trumpet (Fig. 35.2). The roots form thick tubers rich in starch. The species is native to tropical America. Indonesia is becoming the second centre of diversity.
It has many varieties distinguishable by the size, shape and taste of tubers. There are differences in leaf and stem characters also. It grows well in the lowlands as well as in highlands up to 1500 m altitude. In Java it is much cultivated in the upland fields as well as in the kitchen gardens for direct consumption or for commercial purposes. The stems are often planted for fences bordering a garden or a plantation.
The propagation of cassava is done by cuttings. Harvesting is done 8-12 months after planting. Usually a large plantation uses high yielding varieties without paying any attention of the taste. The young leaves are eaten after being cooked. The tuber is used for making various kinds of food, and in many areas it serves as a staple food. Its protein value is low.
Cassava is monoecious and predominantly outcrossing is mediated by protogyny which leads to high degree of heterozygosity in plants and among populations produced from botanical seed. Cassava varieties are heterozygous individuals which are propagated vegetatively to reproduce the genotype.
The cultivated germplasm has erratic flowering habit and apical dominance, generally producing a single woody stem with 2-3 levels of lateral branching. Although some branching is useful for increasing the leaf area index, too much branching conflicts with the need for uniform vegetative states for propagation and may lead to a lower root yield as a consequence of enhanced interplant competition.
3. Breeding Objectives of Cassava:
a. High yield (>35 t/ha fresh root)
b. High starch (> 25%)
c. High harvest index
d. Responsive to additional inputs
e. Un-branching or late branching plant type
f. Low HCN content
g. Good cooking and eating quality
h. Early harvest-ability
i. Better root storage quality
j. Shade tolerance for use as an intercrop under coconut etc.
k. Wide adaptation
l. Compact branches
m. Compact root system
n. Resistance to major diseases (cassava bacterial blight, anthracnose, brown leaf spot, cassava mosaic virus)
o. Tolerance to adverse soil and climatic conditions
4. Hybridization of Cassava:
This is based on Kawano (1980). Cassava is monoecious. On a given branch, female flowers always open first and the male flowers follow after 1 or 2 weeks. Both self and cross-pollinations occur naturally in cassava.
There seems to be no physiological or genetic mechanism to prevent self-fertilization in normally flowering types. No cross-incompatibility has been found so far. The pollen are relatively large in size and sticky, therefore, natural pollination by wind is unlikely. Several species of wasps and bees are the main pollinators.
Self-pollination varies between 0 and 100%, depending upon the genotypes and planting distance. A distance of 30 m between pure stands of different genotypes is sufficient to prevent cross-pollination between two populations. However, 500 m is suggested for a perfect isolation of two populations in genetic studies.
Large cloth bags measuring 20 by 25 cm are used to protect flowers from undesired pollen. Small cloth bags measuring 10 by 18 cm are used to catch matured seeds. The bags can be made of any cotton cloth of 60 to 80 mesh, and should have a string attached to close the mouth of the bag on the plant.
Tags of 3 by 4 cm are used to identify pollinated flowers. Bottles of 3 to 5 cm in diameter and 5 to 8 cm high are used to transport male flowers during pollination.
After a little experience, the breeder can determine with relative ease in the morning which particular female flower will open that day. The surest way to distinguish them is to open one petal of on unopened female flower in the morning. If a drop of nectar is seen on the basal part of the pistil, the flower will open in the afternoon of the same day.
Emasculation is not needed because female and male flowers are separate. Male flowers open 1 to 2 weeks after the female flowers opened within the same inflorescence. By the time male flower open, the female flowers of the same inflorescence have developed into fruits or have died.
Female and male flowers usually begin to open from 1200 to 1400 hours and remain open about 1 day. To prevent stray pollinations, the flower branches are covered by large cloth bags which also identify the female flowers to be pollinated during the day.
Recently opened male flowers are picked off the branch during the first hours of the afternoon and carried in small bottles. Pollination can be done most efficiently between 1300 and 1700 hours. Because both female and male flowers are large and the pollen is sticky, pollination is easy and requires no special tool. One male flower can be used for the pollination of three female flowers.
After pollination, the female flowers can be covered with a large bag or left exposed. Success rate is higher if the female flowers are left exposed after pollination in some genotypes.
The possibility of hybridization by stray pollen seems to be very low once the stigma is pollinated, even if the female flowers are left exposed after pollination. Each flower branch on which the pollination is made is marked with a tag indicating the cross combination, date of pollination, and number of female flowers pollinated.
If the female flowers are exposed after pollination, they must be covered with a small bag 1 or 2 weeks after pollination. By this time the pollinated female flowers will have developed into young fruits. The bag protects young fruits from fruit fly attack and catches mature seeds which fall off naturally about 3 months after pollination.
One female flower can produce up to three seeds, however, it is difficult to obtain an average of 2.0 seeds per female flower. Cassava is a highly heterozygous species and has extremely high inbreeding depression. After one cycle of selfing, some plants are so weak that they cannot produce enough male and female flowers for further hybridization. Thus, selfing is not a major part of hybridization programs in cassava.
5. Breeding Methods of Cassava:
The breeding methods at IITA/CIAT have generally been as follows:
Domestication of cassava probably began with selection for large roots, more erect plant type with less branched growth and the ability to establish easily from stem cuttings. Mixed cropping systems in which several genotypes are grown in the same field probably provided opportunity for new variability to occur in sexual seed, upon which selection was practiced.
Selection worked towards virtual discontinuation of seed propagation and tended to encourage sterility and/or dormancy in sexual propagation. Breeding objectives for cassava have distinct regional differences that are largely determined by biotic or abiotic constraints.
Resistance to cassava bacterial blight (Xanthomonas campestris pv. manihot), root rots (Phytophthora spp., Diplodia spp. and others), white flies, mites and thrips are under selection in breeding programmes in Latin America, as are root dry matter content, cyanogen content and several features of root and plant appearance.
The recognition of variability in vitamin A content, associated with yellow root parenchyma, has led to efforts to develop high-vitamin gene-pools. Post-harvest shelf-life and the functional properties of starch in processing are also in early stages of investigation at both CIAT and IITA.
Useful variation in tolerance to water and nutrient-limited conditions, including acid, low-phosphorus and high-aluminium soils, has been found in cultivated germplasm and is being incorporated into improved germplasm.
Since the 1970s, the genetic base for cassava improvement at IITA has consisted of local cultivars from within Nigeria, selected clones from crosses originally made in East Africa between cassava (M. esculenta) and its related wild species M. glaziovii and segregating progenies from East Africa, Latin America and Asia.
The most pressing early objectives of IITA’s programme were to breed high-yielding varieties with resistance to African cassava mosaic disease (ACMD) and cassava bacterial blight (CBB).
Other breeding objectives include high-yielding clones with yellow root flesh that have high carotene content and low cyanogenic potential (CNP); good gari (grated, fermented and roasted fresh cassava) quality: resistance to cassava mealy bug (CM), cassava green mites (CGM) and cassava anthracnose disease (CAD).
African national breeding programmes are also concerned with resistance to the devastating ACMD. Which is presently threatening survival of the crop and of genetic diversity in severe epidemics.
A systematic and agro-ecologically based germplasm introduction scheme to broaden the genetic base of cassava in Africa was initiated by IITA and CIAT in 1990. Crosses are made at CIAT, Colombia, between parents of Latin American origin that are adapted to four different agro-ecologies of South America with homologues in African and improved ACMD- resistant clones from IITA.
By 1994, more than 200,000 botanical hybrid seeds comprising over 1000 families were introduced for evaluation in Nigeria through IITA, observing strict quarantine regulations for intercontinental exchange of germplasm.
Sources of germplasm for cassava improvement in Asia have been local landraces for direct release, selections from open-pollinated seed originating in germplasm collections and controlled inter-varietal hybrids, as well as crosses between local and introduced (CIAT) clones.
Indian Cassava Mosaic Virus, related to African Cassava Mosaic Virus, is a threat to production which does not at present affect cassava in Latin America. As disease and pest pressure on cassava is low in Asia, primary objectives are to increase productivity and starch content.
Though cassava has been under cultivation in India for almost one and a half centuries, systematic research of this crop was lacking till about 1942 when certain research projects were launched by the Central Research Institute at the then Travancore University in Trivandrum.
In 1951 the work was considerably expanded under a scheme of research jointly financed by the Indian Council of Agricultural Research and the Government of Travancore-Cochin. In 1963 the Indian Council of Agricultural Research established the Central Tuber Crops Research Institute (CTCRI) at Trivandrum (now Thiruvananthapuram) to intensify the research on root and tuber crops.
The significant activities in cassava breeding at this institute include the following:
Germplasm Collection and Evaluation:
Considerable amount of variability exists in this crop in India and much more is available in other cassava growing regions, like tropical South America which is the centre of origin of cassava. Presently the number of cassava germplasm accessions in CTCRI is about 1400. including 60 hybrids, seven wild species, 605 indigenous cultivars and 689 exotic collections the latter received from CIAT, Brazil, I1TA, Madagascar. Malaysia, Sri Lanka, etc.
The evaluation of the early introductions resulted in the identification of two promising cultivars, M4 and M-6. both from Malaysia. Accessions S-1309, S-1310, S—1315, S-2407 and S—2331 were reported to be high yielding.
An indigenous selection, S-856, a high yielding cultivar with early harvest-ability, was released in 1987 under the name Sree Prakash. The yield potential and other features of this and other cultivars are given in Table 35.1.
Seventy five clones selected from true seeds received from CIAT were evaluated along with Sree Visakham and Sree Sahya as controls. Most of these were highly branching types and the root yield per plant ranged from 0.6 to 6.0 kg. Selections from CM 3433, CM 3963 and CM 4008 produced higher yields than the controls.
A total of 654 accessions of indigenous and exotic origin were screened for carotene, 21 clones showed variable intensities of yellow flesh and had a β-carotene content, which ranged from 65 IU in CI-507 to 670 1U (0.40 mg/100 g) in CE-373.
Four selected clones were grown in isolation and the first cycle population was harvested. In this population, 43 plants had roots with different intensities of yellow colour, of which 19 were yellow to deep yellow, and the carotene content ranged from 540 to 1500 1U.
Of the 1214 accessions screened for flowering for two seasons, only 692 flowered. Different types of pollen sterility were noted and categorized. Among these, 20 indigenous and 13 exotic accessions were completely male sterile. Ambakadan, an indigenous accession, had a spectacularly high yield of 51 kg/plant when grown at CTCR1. The plants were highly branching and the roots were very long with good culinary qualities.
Studies on developing new cultivars with shorter crop duration, higher yield and resistance to virus are also conducted at Tamil Nadu Agricultural University. From this programme two improved cultivars, i.e. Co-1 Tapioca and Co-2 Tapioca have been released. Both have a harvest-ability of 8.5 to 9 months, they have high yields of 35 to 38 t/ha, high starch content of 35%.
The hybrid ME-120 has been found promising, exhibiting field resistance to mosaic virus with yields up to 45 t/ha and a root starch content up to 38.6%. Ichapuram Local is a promising cultivar of Andhra Pradesh. Inter-varietal Hybridization
The wide genetic variability in cassava can be judiciously used for the development of agronomically better cultivars by extensive hybridization and selection of plants with desired combinations of characters. Being highly heterozygous, such crosses can be expected to show wide genetic variation, facilitating selection in the first seedling generation.
Inter-varietal hybridization was attempted as early as 1942, resulting in high yielding cultivars like H-96 and H-105. An extensive inter-varietal hybridization programme, launched by CTCRI in Trivandrum, culminated in the development of promising hybrids like H-97, H-86, H-165, H-57, H-75, H-201, H-220, H-226, H-1687 and H-2304.
Of these, H-57, H-165 and H-226 were released for general cultivation in 1971, while H-1687 and H-2304 were released in 1977 under the names of Sree Visakhain and Sree Sahya, respectively. The pedigree yield potential and other desirable characters of the released hybrids are give in Table 35.1.
All the above mentioned hybrid selections have a high yield potential, but in cooking quality they are not as good as M-4, which has a high acceptability among the general public for its excellent culinary quality.
Hence, it was felt imperative to improve the culinary quality of the existing high-yielding cultivars by incorporating in them the good culinary quality of cultivars like M-4, Kalikalan, etc. Hence several crosses were made during 1986 and 1987 among H-165, H-226, H-1687 and H-2304, as well as between these cultivars and those with good culinary quality such as M-4 and Kalikalan.
Seven hybrid clones, i.e. 21/86, 23/86 (H-165 x M-4), 43/87 (H-165 x H-2304), 15/86, 16/86 and 11/86 (H-2304 x M-4) and 50/86 (H-1687 X H-2304) were significantly higher yielding than M-4 and statistically equal to H-2304, the highest yielding control.
Another promising hybrid (H-22-86) from Sree Prakash X M 4 combined high yield potential, early harvest-ability and good culinary qualities. It was tolerant to Cercospora and susceptible to cassava mosaic disease (CMD).
Heterosis Breeding:
Cassava is highly heterozygous and this heterozygosity is perpetuated through years of asexual propagation. Through selfing, good inbred lines could be developed and promising hybrids could be obtained by crossing of superior inbreds.
A project on these lines was initiated at CTCRI in 1981. Marked inbreeding depression was noted in S1, S2, S3 and S4 generations for root characters, shoot weight, total biomass and harvest index. Six selected S4 lines were crossed in a diallel fashion and hybrids were reported to be in the seedling stage.
Polyploidy:
Induced tetraploids by colchicine treatment have been found to be poor yielding and less adapted to normal field conditions. However, the triploids, developed by crossing tetraploids and diploids, have been reported to be promising and it therefore seems possible to improve cassava by producing new chromosomal lines, in which the chromosome number does not go beyond an optimum level.
To produce triploids, a number of tetraploid clones are crossed with diploids. However, successful production of triploids in large enough numbers to be able to exert selection pressure, is possible only in a limited number of combinations, probably due to the operation of cytological diploidization over a period of time in some tetraploids.
The triploid plants derived from OP-4 (2x) X S-300 (4x) and OP-4 (2x) x H-2304 (4x) consistently produced roots with high DM in the seedling and succeeding clonal generations, which ranged from 34 to 43%. Some of the triploids recorded high starch content from the eighth month onwards, being significantly higher than that of the control. Among these, the triploid 76-9 had a yield similar to that of H-2304, the released cultivar, at CTCRI.
Induced Mutation:
A mutant with short petiole has been reported to have practical application on the basis of research at CTCRI. Successful induction and recovery of more than 50 mutants could be accomplished at CTCRI by single-node cutting propagation of acute gamma-irradiated stakes of M-4 and pruning of the MV 1 plants.
A preliminary evaluation of 19 of these showed that there was considerable variation in root yield, DM and starch content, as well as the HCN and chlorophyll contents.
The photosynthetic rate, measured in terms of ppm CO2 uptake/m2/sec of 15 mutants showed a variation from 16.56 to 97.58, compared to 28.00 in the control during the fifth-month growth stage. Similarly, the proline content, considered to be a major constituent in controlling drought resistance, varied from 16.0 mg/g to 53.0 mg/g compared to 18 mg/g in the control.
Biotechnology:
Techniques for the elimination of cassava mosaic disease, developed first in Canada by Kartha and Gamborg in the year 1978 were modified by scientists to suit the local conditions. Disease- free plants of M4, H-97, H-165, H-226, Sree Visakham, Sree Sahya and Sree Prakash have been produced and multiplied, and are now being distributed to farmers in India.
In field trials these plants out-yielded the normal symptom-free plants to the extent of 10 to 25%, depending on the cultivars. Sero-diagnostic tests showed that the meristem derived plants were free from cassava mosaic disease during the initial stages.
Propagation by True Seeds:
Studies on true seed propagation in cassava have indicated the potential of increasing the propagation rate more than 30 times, as compared to clonal propagation. In a replicated yield trial seedlings of a few promising lines recorded root yields of 30 t/ha.
The seedling population of a few promising parents had DM contents of 34%, which was higher than that of the control cultivar, Sree Visakham. In the seedling population, CMD was negligible, i.e.. only 1.5 to 2.8% whereas in the clonal population it was 14-15%. Even in the seedling progeny of severely CMD infected parents, disease incidence was similarly low.
6. Use of Double Haploids in Cassava Breeding:
The concept has been advocated by H. Cebballos and his colleagues at CIAT, Cali, Colombia. The process starts with the selection of elite clones themselves or after improvement for tolerance to inbreeding following the S2-recurrent selection.
Once the planted material begins flowering, tissue will be taken for the induction of doubled haploidy through tissue culture protocols developed specifically for that purpose in cassava.
Upon the production of DH tissue or embryos, in vitro multiplication of each line will be carried out, to produce at least 10 hardened plants ready for transplantation to the field. This would take place at the end of the second year of activities.
Several DH lines will be produced and the ten plants representing each of them will be planted in a Clonal Evaluation Trial in the proper target environment. Hopefully these trials will involve at least 200 DH lines.
Selection of these lines will be conducted for relevant characteristics with moderate to high heritability: resistances to diseases and/or insects, plant architecture, root dry matter content, root and parenchyma colour, harvest index, etc.
The selection at this stage operates with twice the additive genetic variance expected to be found in the original population under random mating conditions. Therefore, it is expected that large contrasts will be apparent at this stage. Lines surviving to this stage will have, by definition, reduced genetic load compared with the elite lines from which originated.
While the field evaluation is conducted lab analyses can be simultaneously carried out to obtain the molecular fingerprinting of each line. This will allow for further selection of characteristics difficult or impossible to determine from the field trials.
For instance, marker assisted selection for CMD (Cassava Mosaic Disease) could be implemented in Colombia, although the disease is not present in this country. Also genetic distance among the lines could be determined to facilitate the production of hybrids among the surviving DH lines.
It is expected that from the 200 or more DH lines at least 30 will reach this stage. Although it is clear from the literature that genetic distances have failed to explain satisfactorily the heterosis among inbred lines in maize, genetic distances measured through molecular markers can be used at least to orient the crosses that deserve some priority.
This could be justified until an adequate definition of heterotic patterns is eventually reached. Since the parental materials (DH lines) are homozygous just a few seeds per cross are required at this stage. The only justification for obtaining more than one seed would be to accelerate the time required for evaluation with large number of plants representing each hybrid or clone.
With the production of hybrid from the selected DH lines, dominance effects (heterosis) are generated and, because of the breeding scheme proposed, will be fully exploitable by the cassava-breeding projects. Depending on the number of hybrid seed produced the previous stage, the evaluation and selection of hybrids can be conducted in two successive steps or just one growing cycle.
It is assumed that only ten plants from each cross can be obtained from botanical seed, and therefore the evaluation and selection is conducted in two consecutive growing cycles. The first selection is performed on all the hybrids produced and based on the 10 plants representing each hybrid clone.
Because there is no replication, selection will be based only on high-heritability traits, and in the proper target environment, to allow for the pressure from biotic and abiotic limiting factors. The same evaluation plots are used as seed multiplication plots.
The second stage of selection and evaluation is conducted with about 100 plants (i.e. two replications at two locations with 25-plant plots). Only hybrids that survived the selection process the previous year will be included in this evaluation.
Low-heritability traits are incorporated as selection criteria at this stage. Only a few clones will survive this selection and they will be included in Regional Trials for their eventual release as has been traditionally done.
While the evaluation of hybrids is conducted, their parental DH lines will be planted in the field in such a way that they are about six months old when the results of the hybrid trials become available. As soon as the hybrid trials yield results regarding the best DH progenitors, they will be crossed to generate new genetic material for the following cycle of selection.
The only purpose of these crosses will be to generate F, plants from which to extract flower tissues for the production of a new generation of DH lines. Hybrid trials will not only generate elite clones to be included in Regional Trials and eventually be released as new varieties, but also provide important information about the DH lines that generated the hybrid clones.
This information will be used to determine lines with good general combining ability (i.e. that generate progenies with performances that are better than the mean of all the hybrids evaluated) as well as detecting heterotic patterns. This information is fundamental for deciding the kind of crosses that will be made for the next selection cycle.
This scheme has following advantages:
The emphasis will shift from producing vast number of hybrids hoping that one (or few) will be genetically superior, towards the production of parental lines that will allow ‘to design’ outstanding hybrids in a gradual, consistent and reliable fashion.
Genetic loads will be quickly reduced in elite cassava populations. Hybrids produced from inbred lines will be better than hybrids produced from non-inbred progenitors because genetic load is reduced and because the system allows building up dominance effects.
Germplasm exchange will be greatly facilitated (botanical seed of outstanding parents) with obvious advantages for the cassava research community. Gene exchange will also be greatly facilitated (currently it is very difficult to transfer one valuable gene from its source into an agronomically superior clone: the availability of inbred lines would make the back-cross scheme feasible for cassava).
Inbred materials are genetically stable, they allow the breeder to capture and efficiently exploit the genetic superiority contained in them, therefore, guaranteeing a sustainable and consistent genetic progress that cannot be observed nowadays.
Once a given combination of inbred lines is found (good performing hybrids) the same genotype could be produced at first using botanical seed, and from there by vegetative means. This implies not only a faster multiplication rate but also cleaner genetic stocks (from the phytosanitary point of view). The system allows for the identification of useful recessive traits.
7. Germplasm Conservation of Cassava:
Plant germplasm is a nonrenewable natural resource that is essential to sustain human life. Without counter measures, the diversity of germplasm will be reduced as economic and technological development increases with world population.
Population pressures have led to deforestation, urbanization, industrialization and changes in agricultural methods, all leading to genetic erosion. It is essential to assemble and preserve as much plant genetic diversity as possible. World cassava accessions are maintained at IITA, Ibadan, Nigeria and CIAT, Cali, Colombia.
The cassava germplasm preservation is done under two systems:
1. Ex-situ in field
2. In-vitro in lab
The accessions in field gene bank are regenerated at various intervals depending upon the longevity of given accession. For in-vitro preservation, a small (0.5 cm) part of cassava plant (shoot or nodal axil cuttings) is sealed in a small tube along with a special maintenance medium at 23-25°C temperature 12-hr day/night photoperiod provided with 1000-1500 lux illumination, slightly modified basal Moorashige Skoog (MS) medium, 25 X 150 mm test tubes capped with aluminum foil and firmly sealed with appropriate wrap. Five tubes are maintained/clone.
The cassava segment slowly continues to develop and grow. Each accession in the in-vitro part of collection is maintained individually. As with the field preservation, the accessions preserved in vitro must be regenerated or sub-cultured at various intervals (10-19 months) depending on the cultivars.
The total cost of in-vitro preservation of cassava germplasm at CIAT has been found to be about 53% more than that for field preservation. With respect to cost of maintaining cassava germplasm in field vs. in vitro, the contrary reports are also available. The number of cassava accessions maintained at CIAT are listed in Table 35.2.
