Conservation and Breeding of Ornamental Lathyrus Species

A considerable number of diverse ornamental forms of Lathyrus odoratus exist which require careful conservation if they are not to be lost. Other ornamental Lathyrus species may have potential for increasing their diversity. A seedbank has been established in the United Kingdom to conserve this diversity and make the material readily available to researchers and growers throughout the world. Conventional hybridisation provides an opportunity for the existing diversity in L. odoratus to be extended. Opportunities also exist to create improved decorative forms and new colours in some other species through intra-specifi c hybridisation. Inter-specifi c hybridisation within the genus is in its infancy but early results suggest exciting improvements might be introduced by this method. Such work might also provide benefi ts for comparative genomics in the tribe Fabeae.


Historical Overview
Although Lathyrus species have been grown since ancient times, their cultivation for ornamental purposes can be traced to the sending of the sweet pea, Lathyrus odoratus, from Sicily to England and the Netherlands in 1699.This species quickly became widely distributed for its scent and decorative fl owers.The typical species has maroon and violet bicoloured fl owers but within 20 years a pink and white bicolour cultivar and a white cultivar existed.By 1800, six colours were known and continued to steadily increase but there is no indication that intentional crossing of cultivars was occurring before 1860.Such novel fl ower colours have led to a perception that L. odoratus has a high mutation rate but there is no evidence to support this.
The few decades from 1860 onwards saw several plant breeders actively crossing cultivars to produce new ones, greatly widening the colour range and improving the size and vigour of earlier colours.One of these, Henry Eckford, coined the term 'Grandifl ora' for his Sweet Peas to promote their improved size compared with earlier material.By 1900, dozens of new cultivars were available (Fig. 1).In addition, other Lathyrus species were being grown as part of a wider interest in discovering the world's fl ora.In 1900, a new form of L. odoratus was seen that caused a sensation.It was a cultivar with larger, frilly petals and a longer raceme, making it much more attractive to fl orists than the smaller, plain petals of the Grandifl ora cultivars.This fi rst large, frilly cultivar was named 'Countess Spencer' and all subsequent cultivars with this fl ower form have been known as Spencer sweet peas (Parsons 2000).This new fl ower form, together with the discovery of Mendel's research, led to a frenzy of breeding new sweet peas and a considerable increase in growing sweet peas for cut fl owers and by amateur gardeners.By 1910, California alone was growing around 400 ha of sweet pea seeds and had become the main centre of seed production (Wright 1912).The 20th century saw considerable refi nement of Spencer sweet peas by plant breeders to improve fl ower size, purity of colour, raceme length, and vigour.They also introduced new types having early fl owering characteristics, dwarf and intermediate height, non-tendril (multijugate) leaves, and additional petals.As a result of all this work, the vigour of plants and associated raceme length and fl ower size all increased (Parsons 2011).

Conservation of Lathyrus Species and Cultivars
Cultivation of L. odoratus cultivars has taken place in many parts of the world and remains signifi cant.Perhaps the main activity is in cut fl ower production, often to meet local markets.For example, most northern European countries have at least one producer of cut fl ower sweet peas.Although less popular in the UK nowadays because of their short vase life, they remain popular as a winter cut fl ower crop in Japan.By contrast, few sweet peas are grown by amateur gardeners in Japan but they remain a popular fl ower with UK gardeners.Seed production has become more diversifi ed, with New Zealand rising as an important secondary centre of seed production (Parsons 2009).
A problem of fi eld scale production of L. odoratus seed is that the Spencer fl ower form is recessive to normal form.Records of petal size and plant vigour show that Grandifl ora cultivars, with normal fl ower form, have not changed signifi cantly during the last 100 years of seed generations but Spencer cultivars are frequently seen to revert to clamped keels, either with normal fl ower form and size or with an intermediate size and form.These intermediate fl owers have a clamped keel and frilly petals and have been called Semi-Grandifl ora type (Parsons 2011) (Fig. 2).When cropping a cultivar on a fi eld scale, it is easy to remove any plants with the wrong fl ower colour but reversion in fl ower form, or other faults that arise, are easily overlooked so that commercial stocks of a cultivar deteriorate relatively quickly (Parsons 2011).This problem is made worse because there is no registration authority for Lathyrus cultivars and they are introduced into cultivation without the rigour that applies, for example, to fi eld pea (Pisum sativum) cultivars.Commendable effort has gone into conserving the diversity of forage and grain Lathyrus but there was no readily accessible seedbank conserving ornamental Lathyrus.In the UK and elsewhere, sweet pea commercialisation has been characterised by a relatively short cycle of cultivar introduction, deterioration over several generations and replacement by a new cultivar with the same fl ower colour.Such new cultivars may fi ll the same niche in the marketplace for fl ower colour but each cultivar has its own characteristics, for example adaptation to microclimate and different soils.The fi rst-named author became concerned that cultivars were being lost and that the genepool was being reduced.He had been conserving other Lathyrus species since 1991, but in 1998 established a seedbank to conserve L. odoratus cultivars.Fortunately, Lathyrus seed stores very easily when placed in an air-tight container in an ordinary domestic freezer (Hammett 1979).
In 2005, this collection was recognised as a UK National Collection, trading as Roger Parsons Sweet Peas.It now consists of around 900 L. odoratus cultivars and around 100 taxa from other Lathyrus species.New accessions from around the world are always welcome.Over 100 stocks of annual Lathyrus, mostly L. odoratus cultivars, are grown each year.These are grown in very small quantities, usually between 10 and 60 plants of each cultivar, so that each plant can be assessed for being true to its original qualities.Unlike fi eld cropping, the emphasis is on the quality of the seed stock rather than the quantity of seed produced.It might be thought that fi eld-scale seed producers should do more to maintain the quality of their seed stocks but such neglect has created a niche market for Roger Parsons Sweet Peas.At the end of each season, fresh seed is returned to the seedbank and the surplus is sold.Customers include seed producers who want a true or improved stock of a cultivar that they can then send to be bulked up for fi eld production.Roger Parsons Sweet Peas also retails direct to amateur gardeners since this provides a higher profi t.A limited selection of cultivars and species is advertised on the website www.rpsweetpeas.co.uk and a mail order catalogue is sent each year to previous UK customers.This provides an opportunity to offer people species or colours that they were unaware existed and to encourage them to try something that is currently not popular.However, all this work is a personal domestic effort which makes it vulnerable to being lost.

Opportunities for Lathyrus Breeding
The seedbank provides access to perhaps the world's most diverse collection of material for this genus.Not content with conserving the fullest diversity, plant breeding provides an opportunity to increase the diversity.This may be from recognising the qualities of mutations that arise casually, but more importantly a small programme of hybridising takes place.Up to now, this has consisted mainly of introducing new colours into the late Spencer type that are so popular with UK gardeners, or improving the qualities of existing colours.In 2010, work began on breeding programmes for dwarf sweet peas and for early-fl owering sweet peas.
This work is possible because the fl owers of L. odoratus cultivars have self-pollinated by the time they are open.This allows cultivars to be grown alongside each other yet remain true from seed.There is a very small risk of some cross-pollination occurring which is why rigorous reselection of small batches of a cultivar is the best way to conserve it.In order to hybridise two cultivars, the petals of the seed-bearing fl ower must be unfurled while still in bud at a stage when the gynoecium is suffi ciently mature to receive pollen but the anthers have not yet produced pollen (Fig. 3).The anthers are removed and then pollen from the pollen-bearing parent is applied to the stigma.The petals are then closed and the fl ower is allowed to develop normally, having of course been carefully labelled.Subsequent generations are grown and allowed to self-pollinate until single plant selections result in a seed stock that breeds true for the desired characters (Parsons 2011).

Fig. 3. A sweet pea fl ower ready for hybridization Slika 3. Cvet mirisnog grahora spreman za ukrštanje
There appears to be considerable potential for other people to engage in breeding new ornamental Lathyrus.There are several types of sweet pea that are not receiving the attention they merit.Very little work is being carried out on earlyfl owering cultivars in the UK, which are not as popular there as the summer varieties.However, they have great popularity in other parts of the world and with commercial cut fl ower growers.Japanese cut fl ower growers are quite active with these but are reluctant to allow their cultivars to become available to their commercial competitors.In addition, there is very limited availability of summer-fl owering Multifl ora cultivars and, to the best of our knowledge, nobody is working on these.Some recent interest has been shown in improving dwarf and intermediate height cultivars, which are most popular in the USA, but there is considerable scope for the plant breeder to work on these.
The everlasting pea, L. latifolius, is a popular garden plant in the UK and is grown as a cut fl ower crop in some countries.Another perennial garden species is the dwarf spring pea, L. vernus.Both of these have a small number of colours that have arisen through mutation and it seems that their position refl ects that of L. odoratus in 1800.However it is not simply the occurrence of mutations that has led to such diversity in L. odoratus but human intervention is needed to identify and secure desirable mutations.It may be that by conventional intra-specifi c hybridisation, other Lathyrus species can be induced to provide a broader range of colours and that other qualities can be improved, for example fl ower size in L. vernus (Fig. 4).New colours are arising in L. odoratus partly as a result of inter-specifi c hybridisation.Lathyrus species cannot easily be induced to hybridise but cultivars have now become available that arise from a cross between L. belinensis and L. odoratus, using embryo rescue (Hammett et al. 1994).Hybrid material tends to adopt the morphological features of one parent or another after a few generations but may still introduce novel qualities.Material from this particular cross is still being evaluated but the potential for resistance to powdery mildew (Erysiphe polygoni) has already been reported (Poulter et al. 2003).Personal observations suggest extended vase life may be possible.

Conclusions
A considerable number of diverse ornamental forms of L. odoratus exist which require careful conservation if they are not to be lost.Other ornamental Lathyrus species may have potential for increasing their diversity.A seedbank has been established in the UK to conserve this diversity and make the material readily available to researchers and growers throughout the world.Conventional hybridisation provides an opportunity for the existing diversity in L. odoratus to be extended.Opportunities also exist to create improved decorative forms and new colours in some other species through intra-specifi c hybridisation.Inter-specifi c hybridisation within the genus is in its infancy but early results suggest exciting improvements might be introduced by this method.Such work might also provide benefi ts for comparative genomics in the tribe Fabeae.