The genus Lycoris (Amaryllidaceae) consists of about 20 species, all of which are
confined to temperate China, Japan and Korea.
Cytological investigations, including a reexamination of the karyotypes of 14 taxa, measure-
ments of relative nuclear DNA content, and meiotic configuration observations on some specific
forms and interspecific hybrids, have been carried out by the present authors in order to re-eva-
luate the mode of karyotype evolution and the role of hybridization in the speciation of Lycoris.
These have resulted in a new theory for explaining the karyotype evolution in the genus, which
will be considered elsewhere. The present paper deals with observations on karyotypes of 11
species, 1 variety and 2 artificial hybrids.
Results obtained through karyotype analysis, as shown by the data in Table 1, Plates I-VI
and Figs. 1-2, reveal that: (1) the karyotypes of Lycoris rosea, L. radiata var. pumila, L. spren-
geri, L. haywardii, L. caldwellii, L. squamigera and L. radiata are, on the whole, consistent with
those reported by the previous authors[1,2,3,4,5,8,10,12];(2) the I (rodshaped) chromosomes of L.
chinensis and L. longituba are all T’s (telocentric) instead of t’s (acrocentric) or t(Sat)’s;
(3) the three materials of L. aurea of different sources have shown a karyotypic differentiation:
one with 2n=14=8m+6T, and the others with 2n=16=6m+10T: (4) both of the karyotypes
of L. straminea and L. albiflora are 2n=19=3V+6I, inconsistent with 2n=16=6V+10I for the
former and with 2n=17=5V+12I for the latter as reported by Inariyama (1953), Bose and
Flory (1963) and Kurita (1987).
The following aspects are worthwhile discussing:
1. The types of chromosomes.
Karyotype analyses reveal the existence of three major chromosome types in Lycoris: (1)
m (metacentric) chromosomes: (2) t (acrocentric) chromosomes, with short arms, (3) T (te-
locentric) chromosomes, sometimes with dot-like terminal centromeres. To distinghish t’s from
T’s is of paramount importance for solving the problem of karyotype evolution in Lycoris.
Bose (1963) pointed out that in the species with 2n=22, all I chromosomes were t’s, while in
species with 2n=12-16, all I chromosomes were T’s. Our results of chromosome observations
are consistent with Bose’s remarks. Some authorst[3,6] have probably mistaken the dot-like ter-
minal centromeres of T’s of L. longituba and L. chinensis as the short arms of t’s.
2. The significance of Robertsonian change in karyotype evolution.
Although chromosome numbers and karyotypes are very variable in Lycoris, as shown in
Table 1, the total number of arms of a chromosome complement of any species is always mul-
tiples of 11. Hence, it seems likely that Robertsonian changes have taken part in karyotype
alteration, The genus has a series of basic chromosome numbers: 6, 7, 8 and 11. But which is
the most primitive one? It is uncertain whether a successive decrease in chromosome numbers
as a result of Robertsonian fusion or a gradual increase in chromosome numbers brought about
by fission (fragmentation) has been the essential mechanism for karyotype evolution and spe-
ciation in Lycoris. These problems are of crucial importance and will be discussed in our
subsequent papers.
3. The origin of polyploids.
As evident from Table 1, there are two levels of ploidy differentiation in Lycoris: (1) di-
ploids with 2n=22 or the equivalent of 22, (2) triploids with 2n=33 or the equivalent of 33.
The most common way of origination of triploids in plants is the hybridization of diploids with
Tetraploids. But tetraploids have never been found in Lycoris. Thus, it is suggested that the
triploids have originated from the combination of an unreduced gamete of a diploid with a
normal gamete of another diploid.
4. The role of hybridization in speciation.
Results of karyotype analyses show that hybridization has taken an important part in the
speciation of Lycoris. Two types of hybrids have been found: (1) 2n=19= 3V+ 16I, L. strami-
nea, L. albiflora and the two artificial hybrids L. sprengeri×L. chinensis and L. haywardii×
L. chinensis all possess this karyotype. It could be seen from the above chromosome number and
karyotype that this sort of karyotype is exactly half of the total sum of 2n=22I and 2n=16=
6V+10I. It is, therefore, quite evident that taxa possessing this karyotype are all diploid
hybrids of 2n=22 and 2n=16, (2) 2n=27=6V+21I, L. caldwellii and L. squamigera possess
this karyotype. It is reasonable to assume, too, that they are segmental allotriploids and have
arisen from the combination of an unreduced diploid gamete of 2n=16 and a normal haploid
gamete of 2n=22. The origin of the hybrid karyotype 2n=17=5V+12I reported by Inari-
yama (1953) is similar to that of 2n=19, except that one of the parents possesses 2n=12=
10V+2I instead of 2n=16=6V+10I. The origin of the other hybrid karyotype 2n=30=3V+
27I reported by Bose (1963) is similar to that of 2n=27, but the diploid gamete comes from
taxa possessing 2n=22 instead of 2n=16.
