Breeding of Saffron (Crocus sativus): Possibilities and Problems
Y.M. Agayev, A.M. Shakib, S. Soheilivand, M. Fathi
Agricultural Biotechnology Research Institute of Iran (ABRII)
P.O. Box 31535-1897
Keywords: clonal selection, chromosome doubling, hybridization, unusual methods,
Breeding of saffron (Crocus sativus L.), its position, urgency and topicality
of the problem are considered in the present review. New unusual strategies are
proposed for genetic improvement of saffron in order to increase yield
production both in quality and quantity. Complexity and extraordinary feature
of the problem is drawn. The ways of breeding are analyzed and recommended
in close acquaintance with exceptionally specific biological peculiarities of this
triploid plant. The application of contemporary biotechnological methods makes
solving this problem more feasible. The present report is a reflection of a large
research project related to the problem of saffron breeding, started in ABRII in
2003. Underlying reason of the mentioned project is our firm conviction based on
many years of studying the biology, systematics, cytogenetics, and embryology of
saffron that in spite of intricacy of the subject, not only breeding (especially
clonal selection) of this plant is possible, but also is necessary and this problem
for today stands rather burning. It promises to give annual large profit, where
the plant is cultivated in a vast expanse.
It is well known that saffron (C. sativus L.) is a very valuable irreplaceable
spice with exceptional medicinal properties which has been cultivated since ancient
times. There are no wild forms and it exists solely in culture. Being triploid with
chromosome number 2n=3x=24 and basic number of x=8, saffron never bears seed
and it is propagated exclusively in a vegetative way by corms. Its vegetation period
starts in autumn, continues through winter and finishes in the middle of spring.
Therefore its growth and development is very slow, facing the cold period of the year
which results in low productivity of the plant. However, the demands of world
industry for saffron product are very high. Among countries cultivating saffron, Iran
is at the first place. More than 90 % of the world production falls onto Iranian saffron
which has a great importance in the economy of the country. The area under saffron
cultivation in Iran reaches about 80 000 hectares with an annual production of about
250 tons of the dry stigmas. In recent years this amount has been remarkably
increased, which was achieved mainly by extension of the planting areas but not due
to the increased yield capacity in the area unit. Beside other factors new high yielding
cultivars of saffron are required to solve the problem. It seems that the genetic
improvement of saffron and creation of new high yielding cultivars in the past was
impossible owing to the complexity of the problem. Only just the traditional methods
of breeding are not promising here.
The literature on saffron breeding is very limited. There are known old works
from Azerbaijan Republic more or less concerning with the breeding problems
(Rzakuliyev, 1948, 1959; Kapinos, 1965; Muzaferova, 1970; Agayev et al., 1975). A
lot of work has been carried out using tissue culture. (Homes et al., 1987; Dhar et al.,
1988; Chichiricco, 1999; Munshi and Zargar, 1991; Munshi, 1992; Dhar and Sapro,
1993; Piqueras et al., 1999). Ascertaining the specified activities at the same time it
should be admitted that for today on arena there is only one cultivar of saffron.
Our attitude to the given problem differs. The urgency of saffron breeding
problems and the necessity to solve them with the application of new extraordinary
approaches was stated before (Agayev, 1994a,b). Current work on the breeding of
saffron was conceived and planned up to the smallest details by the conducting
authors and submitted as a new project with 15 subprojects. The project with 9
subprojects has been accepted to be performed since the autumn of 2003 in ABRII. It
is believed that saffron plant being cultivated in different countries under different and
constantly varying land and climatic conditions, has faced with countless different
stress situations during a rather long period of time (thousands years) and undergone
the various mutations. Although these mutations were not redistributed between
plants because of their sterility, but are accumulated in populations and kept till now.
With mutations, small and large, on our view, should be encompassed all attributes,
morphological and physiological, peculiar to saffron plant.
Clonal selection independently and in combination with the experimental
polyploidy and hybridization involving wild close relatives of C. sativus is mostly
promising. Methods of in vitro technique and molecular genetics should be also
applied if necessary.
WHY SAFFRON EXISTS JUST AS ONE CULTIVAR
Ancient history of saffron cultivation goes back to many thousands years ago.
It was originated very likely in Greece and then distributed in the other Mediterranean
and Near East countries (Turkey, Italy, Azerbaijan, Iran, Iraq, North of India, etc.). In
all over the world saffron is known as one cultivar, as descent of certain triploid
sterile plant arisen once spontaneously in nature which was caught by sight of man
and involved into cultivation (Mathew, 1977). It has been propagated and still
continues to be propagated vegetatively. There is a supposition that saffron as a clone
can be scarcely changed genetically and its improvement is hardly possible through
clonal selection (Dhar et al., 1988, Piqueras et al., 1999).
Meanwhile, other suppositions exist as well. For example, Rzakuliyev (1959)
investigating in Apsheron (Baku) specimens of saffron obtained from 6 regions (2
regions in Italy, France, Istanbul , Yalta and Mashtağa) during 3 vegetations in 1934-
1937 concluded that it is possible to create a new high yielding cultivars of this plant
on the basis of clonal selection. Kapinos (1965) studied the morphogenesis and
cytoembryology of Crocus sativus also under climatic conditions of Apsheron and
came to the conclusion that this plant represents variegated blend of genetically
heterogeneous forms – clones, and clonal selection on it would be very promising.
Apparently, the lack of new cultivars of saffron at present may not be
explained by the impossibilities of the improvement of this plant through clonal
selection. To solve this problem, it is necessary for a researcher to be properly
acquainted with the biology and genetics of saffron and work up subtle puzzled
methods of the clonal selection especially for saffron, different from those of the other
Two main difficulties of saffron breeding through clonal selection, in our
opinion, are as follow:
1) Difficulty in the recognition of the clones. In any plantation saffron is represented
by plants existing in highly different “ages” of individuals connected with different
sizes of corms underground. Above the ground, these plants differ in the size and
number of their flowers (at the stage of flowering), also in their number and size of
leaves. If some plants are sharply different from the others in certain characters,
interesting for the aim of breeding, a researcher can not practically identify them.
Naturally he does not see corms underground, and can not elucidate the cause of the
mentioned differences: whether these differences are due to the age (size) of corms, or
because of the genotypes of plants. So, genetically different (if exist) and similar
plants will continue to grow together and not be used as a subject for breeding.
2) Difficulty in the multiplication of clones and bringing them to cultivars. Let us
suppose that farmer recognizes certain plant(s) which could be used as a clone with
good economic characters. Multiplication of such clone(s) would be turned as
insoluble problem. One saffron corm at planting with proper care produces on average
4 corms of middle sizes during vegetation (one year). At such intensity of propagation
it could be brought to about 1000000 corms only after 17 years. This amount could be
enough for planting on the area of 2 ha. It is clear that a farmer will accomplish never
such an exhausting work of many years. Therefore the farmers would not pursue the
aim to make new cultivar of saffron even if they have been lucky to find some clones
with very highly expressed economically valuable characters. Concerning to the
researcher, in our opinion, he could be able to do it in the favorable conditions,
namely in the case of worthy appreciation of his long term hard work on this way by
existing law on breeding. That is why in all over the world saffron has been remained
as one cultivar despite of existence in culture during hundreds and thousands of years.
Alternatively, genetically changed superior plants of saffron could be
propagated rapidly via in vitro technique. Investigations in this direction are very
promising. Unfortunately, the experiments pursuing rapid corm propagation of saffron
have not been successful so far and a few in vitro developed corms had been
produced. Matured corms of in vitro origin in mass production had not been
CONFIDENT STRATEGIES OF CLONAL SELECTION OF SAFFRON
With the object of breeding new cultivars of saffron with economically
valuable traits, two principles can be applied:
a) searching, identification and separation of superior clones in existing plantations,
b) creating new valuable forms experimentally.
Suggesting that in the existing plantations clonal selection of saffron is possible and
promising, we are guided by following considerations. Having an ancient history of
cultivation, saffron apparently should contain a lot of genetically changed forms
(clones) as the result of mutations in somatic cells. The task is to find and study them
individually, to separate the economically valuable forms and to bring them to the
new industrial cultivars. But the matter is how to do it?
In order to recognize the changed forms, all of the corms were separated into
groups according to their weights. Difference between adjacent groups was 1.0 g, or
0.5 g, or even less than 0.5 g. For the first time, we contented the difference of 1.0 g.
The groups were as 3.0-3.9 g, 4.0-4.9 g, 5.0-5.9 g, etc. up to 16.0-16.9 g and more.
Ignoring slight differences between weights of corms within each group (not more
than 1.0 g), we conditionally accepted that they were of the same weight. The corms
of each group were separately planted in lines in soil. In each hole (cluster) only one
corm of a given weight was planted. So we were able to compare plants within each
group separately, and to investigate the existence and kind of changed forms. In this
way, we could recognize within each group changed forms with valuable breeding
characters such as multiflowering, larger flowers, flowers with stronger color and
more aroma stigmas, tall flowers, early or late flowering, simultaneously flowering,
lack of leaves at flowering time, number of leaves, their vigor, etc. These
investigations were reiterated a few years for elucidation of the inheritance of
established changes. At the end of such work, new clones with economically valuable
characters were finally established. Next work will involve operations to propagate
and use them in the breeding programs.
OVERVIEW OF OUR BREEDING EXPERIMENTS ON SAFFRON IN FIELD
AND GREENHOUSE at ABRII, KARAJ AND SOME PRELIMINARY
An experimental field with an area of 3 ha in Kamalabad (near Karaj) was
divided to six parts (plots). In five plots corms from five old saffron cultivation
regions of Iran: Gonabad, Ghaen, Estahban, Ferdows and Birjand were planted
separately and one plot was assigned for nursery. Corms were divided, as it was
explained above, according to their weights into groups (3.0-3.9 g, 4.0-4.9 g, up to
10.0-10.9 g and more) and planted separately in lines in plots. The distance between
adjacent lines was 50 cm; between corms in each line of 3.0-3.9 g and 4.0-4.9 g 12.5
cm, and for corms of 5.0-5.9 g and more 25 cm was considered. In each hole (cluster)
only one corm was planted (Table 1).
In the greenhouse, corms of C. sativus were planted in mixture of soil, sand,
vermiculite (or perlit) and peat moss in pots of 5 liters in capacity and in the plastic
baskets of two different sizes (21x33x21 and 40x59x30 cm). At the end of winter,
intensively growing apical meristems of plants were treated by modified injection
method with the solutions of different concentrations of polyploidizing reagents:
colchicine and reagents of herbicide origin. Plants of 668 clusters were treated with
colchicine, 1180 clusters with oryzalin, 1247 clusters with trifluralin, 1236 clusters
with APM and 844 clusters with propizamide. After treatments plants continued their
vegetation during 1.0-1.5 months, having completely green leaves, and then their
leaves became gradually yellow and dried. At the same way many plants of close
diploid wild relatives of C. sativus, namely, C. cartwrightianus Herb., C. hadriaticus
Herb., C. pallasii Goldb. and C. asumaniae Mathew & Baytop were treated. The
increase of ploidy level of specified plants will allow us to accomplish wide range
crossings and produce desirable hybrids for breeding purposes.
The number of flowers in several lines was counted in Gonabad and Ghaen
plots nearly at the end of flowering time after 2 years of planting saffron corms.
Analyzing the results of this counting showed that within each weight group number
of flowers was differing in large extent in different clusters (holes), despite planting
corms of the same weight in each line and only one corm in each hole. To say roughly
35-50 % of clusters had no flowers at all, 20-25 % had one flower, 13-15 % two, 5-
10 % three, 2-7 % four, 1-2 % five flowers, etc. Unique clusters had 8-10 and even up
to 12-13 flowers. This regularity was similar in Gonabad and Ghaen plots, and no
doubt that the same will take place also in the rest of the three plots. No doubt that the
analogous differences between plants should be observed also in other characters.
This result shows that the analyzed saffron plantations are heterogeneous. They are
consisted of the genotypes different from each other by different characters in large
scale. This regularity has sharp expression in the character referred to as number of
flowers. We do not doubt that discovered multiflowering plants in the next
generations will keep their property and continue to be as multiflowering plants, and
plants with no or few flowers will stay as so as they were registered, considering that
each observed saffron plant is individually a clone. Discovered law governed
regularity opens new wide prospects forward for clonal selection of saffron, C.
Agayev, Y.M. 1994a. Some urgent problems of genetics, cytogenetics and breeding of
saffron. Abstracts of the 2nd Symposium on Saffron and Farming of Medicine
Plants. Gonabad, Iran. P. 12.
Agayev, Y.M. 1994b. Origin of saffron and its karyotype analyses. Abstracts of the
2nd Symposium on Saffron and Farming of Medicine Plants. Gonabad, Iran. P.
Agayev, Y.M. 2002. New features in karyotype structure and origin of saffron Crocus
sativus L. Cytologia 67:245-252.
Agayev, Y.M., Muzaferova, R.S. and Savchenko, S.P. 1975. Results of the
experiments of saffron corm treatments with colchicine solutions. Vestnik
Selskokhozyaystvennoi Nauki (Bulletin of Agricultural Science, Moscow),
Chichiricco, G. 1999. Sterility and perspectives for genetic improvement of Crocus
sativus L. In: Saffron, Crocus sativus. Medicinal and Aromatic Plants – Industrial
Profiles. (Ed. By Negbi, M.). Hardwood Academic Publishers. P. 127-135.
Dhar, A.K., Sapru, R. and Rekha, K. 1988. Studies on saffron in Kashmir. 1.
Variation in natural population and its cytological behavior. Crop Improvement
Homes, J. Legros, M. and Jaziri, M. 1987. In vitro multiplication of Crocus sativus L.
Acta Horticulturae 212:675-676.
Kapinos, G.E. 1965. Biology of development bulbous and cormous plants in
Apsheron. Azerbaijan Nat. Acad. of Sciences, Baku.
Mathew, B. 1977. Crocus sativus and its allies (Iridaceae). Plant Syst.and Evol. 128:
Munshi, A.M. 1992. Genetic variability for important traits in saffron (Crocus
sativus). Crop Research (Hisar) 5: 326-332.
Munshi, A.M. and Zargar, G.H. 1991. Variation in natural population of saffron
(Crocus sativus L.) crop in Kashmir and performance of some selected sub-
population. Phytobreedon 7: 62-67.
Muzaferova, R.S. 1970. Primary results of a study of influences of the mutagenic
factors on changing of saffron. Proceedings of the Institute of Genetics and
Selection of the Academy of Sciences of Azerbaijan SSR, Baku, "Elm". P. 194-
Piqueras, A., Han, B.H., Escribano, J., Rubio C., Hellin, E. and Fernandez J.A. 1999.
Development of cormogenic nodules and microcorms by tissue culture, a new tool
for the multiplication and genetic improvement of saffron. Agronomie 19:603-
Rzakuliyev, I.M. 1948. Flowering biology of saffron (Crocus sativus L.). Izvestiya
AN Azerb.SSR (Baku), 2:105-114.
Rzakuliyev, I.M. 1959. A study of different specimens of saffron in condition of
Apsheron. AGU, Uchyonyye Zapiski, Biol. Ser., 5:3-8.
Table 1. Description of the experimental field in Kamalabad with saffron plantings
originated from the five ancient cultivation regions in Iran
Total Numbers of lines (above) and corms (below) with weights:
Name of weight
plots and Area of
planting (m²) planted 9.0-9.9 g
3.0-3.9 4.0-4.9 5.0-5.9 6.0-6.9 7.0-7.9 8.0-8.9
date corms total and
g g g g g g
Gonabad 124 42 26 28 12 4 4 8
19.11.20 5750 234
03 52164 22690 14728 7540 4460 1600 655 491
Ghaen 65 16 8 20 9 - 6 6
19.11.20 2900 127
03 25479 9392 5026 5909 2596 - 1179 1377
Estahban 89 6 10 16 14 10 7 13
13.10.20 2500 85
04 13525 1950 2880 2600 2100 1525 1090 1680
Ferdows 92 29 18 19 10 5 3 2
20.10.20 2500 92
04 21760 10560 5474 2986 1443 682 352 263
Birjand 106 22 21 22 19 9 5 8
24.10.20 2500 111
05 21712 6050 5550 4493 2858 843 1011 907
476 115 83 105 64 28 25 37
Sum 16150 649 13494
50642 33658 23528 13457 4650 4287 4718