34 RESEARCH REPORTS PNL Volume 12 1980
Jacobsen, E. and W. J. Feenstra University of Groningen, Haren, Netherlands
Pea haploids have not been described in the literature. Because of
technical limitations, only a few practicable methods can be used to induce
and select for haploids. The use of marker genes to select haploids by
parthenogenetical induction is not attractive because the pea is a self-
fertilizing species with a low seed set per fruit. However, anther culture
and selection of parthenogenetical twins are practicable methods.
Selection for haploids by the twin seed method has been practiced in
different crops, such as wheat, rice, and potato. For leguminous crops,
haploids have been selected in this way in yellow lupine (5) and soybean
(1, 2). Since the method of twin seeds has not been mentioned in literature
for pea, we have decided to investigate its potential.
Twenty-three varieties and breeding lines have been studied for the
ability to produce twins. Two thousand or more seeds per line were germi-
nated in sterile vermiculite. After 4-5 days the seedlings were screened.
Screening was carried out using different criteria. The number of cotyledons
served as a reliable criterion for lupine (5), whereas the number of roots
per germinated seed was suitable for potato (4). Pea cultivars 'Diktrom'
and 'Poloma' very infrequently showed the presence of seeds with three coty-
ledons of non-uniform size (Table 1); one diploid shoot always appeared.
Table 1. Lines tested for twin formation showing 3 cotyledons or two
primary roots.
Number tested 20
root 3 cotyledons
St 50 - 74
St 430 - 75
Stehgolt (1978*)
Stehgolt (1976*)
Seeds harvested
It was evident after germination that breeding lines St 50-74, St 430-75,
and cv.'Stehgolt' contained seeds with 2 primary roots. Two different types
of twins were observed: twins with 2 normal-sized shoots and twins with one
large- and one small-sized shoot.
Chromosome numbers of both twin forms have been studied according to
a procedure described by Tjio and Levan(6). Forty-six twins have been studied
in detail and 43 twins have been studied in part. All plantlets contained
the diploid somatic chromosome number. Most twins were fully separated;
possibly all were identical twins.
Because of the relatively high frequency of twin forms in seed of cv.
Stehgolt this variety was studied more intensively. The year of origin proved
to be a very important factor influencing twin frequency. Among 5,000 germi-
nated seeds harvested in 1976 not a single twin was detected, whereas among
10,000 seeds harvested in 1978, 82 twin seeds were found. Investigations
PNL Volume 12 1980
on wheat suggested that polyembryo formation can be caused by herbicides
used at a critical developmental stage (4). For this reason it is not clear
whether the observed variation between varieties in twin formation is based
mainly on genetical differences or on environmental conditions.
The twin method for selection of haploids has been without any success
after examining 76,342 seedlings and investigating 89 twins cytologically.
1. Ahmad, Q. N., E. J. Britten, and D. E. Byth. 1975. J. Hered. 66:327-330.
2. Beeversdorf, W. D. and E. T. Bingham. 1977. Can. J. Gen. & Cyt. 19:283.
3. Frandsen, N. 0. 1968. Z. Pflanzenzuchtg. 59:343-358.
4. Ferguson, J. D., M. M. Mcewan, and K. A. Card. 1979. Physiol. Plantarum
5. Kazimierski, T. and E. M. Kazimierska. 1970. Genetica Polonica 11:62-80.
6. Tjio, J. H. and A. Levan. 1950. An. Estacion exp. Aula. Deiz. 2:21-64.