Genetic Background and Agronomic Value of Leaf Types in Pea ( Pisum sativum )

Pea (Pisum sativum L.) has a compound leaf like many other legume species. The ‘semi-leafless’ pea (afaf TLTL), with all leaflets transformed into tendrils, is considered one of the most important achievements in pea breeding, due to a significantly enhanced standing ability and equally efficient dry matter production in comparison to normal-leafed genotypes (AFAF TLTL). ‘Semi-leafless’ cultivars provide high and stable grain yield and are dominant in the modern dry pea production worldwide. There are also ‘semi-leafless’ cultivars that are autumn-sown and those for forage production. The genotypes with all tendrils transformed into leaflets (AFAF tltl), called ‘acacia’ or ‘tendril-less’, are extremely prone to lodging and may have importance in breeding for forage production. Little is known about the potential agronomic value of ‘acacia-tendril-less’ (afaf tltl) genotypes.


Genetics of Pea Leaf Types
Legumes (Fabaceae, syn.Leguminosae) comprise annual and perennial herbaceous plants, many of which are economically important grain, oilseed and forage crops, as well as shrubs and tropical or subtropical trees.They provide quality protein for humans and animals and enriching the soil by symbiosis with nitrogen-fixing bacteria (Singh et al. 2007).Among the most significant legumes is pea (Pisum sativum L.), a true multifunctional crop that may be used as green forage, forage dry matter, forage meal, silage, haylage, immature grain, mature grain, straw and green manure (Mihailović & Mikić 2010).Pea has also been a subject of genetic investigation since the experiments by Thomas Andrew Knight in the 1790s and Gregor Mendel in the 1860s (Ellis 2007, Ellis 2009).The leaves in most legume species are compound (Lewis et al. 2005).Pea, with its pinnately compound leaf consisting of basal, foliaceous stipules, proximal leaflets and distal tendrils, serves as a model species for understanding compound leaf development (Yaxley et al. 2001, Hofer et al. 2009).It is considered that the role of the class 1 KNOTTED1-like (KNOX1) genes, regulating leaf complexity across vascular plants, in the case of pea and other members of a large subtribe of the Fabaceae is taken over by the gene UNIFOLIATA (UNI), an ortholog of the floral regulators FLORICAULA (FLO) and LEAFY (LFY) from the model species Antirrhinum majus and Arabidopsis thaliana (Gourlay et al. 2000, Champagne et al. 2007).The UNI gene is positively regulated by auxin and gibberellin (Bai & DeMason 2006) and maintains the meristematic potential of the compound leaf, enabling the sequential development of pairs of leaflets and tendrils in acropetal order, as shown by uni null mutants (Fig. 1), with only a single leaflet (Hofer et al. 1997).
Tendrils are characteristic of the tribe Fabeae, to which pea (Pisum L.), vetchling (Lathyrus L.), vetch (Vicia L.), lentil (Lens Mill.) and vavilovia (Vavilovia Fed.) belong, and some species of the close tribe Cicereae.Since nearly all other Papilionoid legumes are untendrilled, it may be assumed that tendrils arose either once, with at least two independent losses, or, twice independently (Hofer et al. 2009).
The size of the stipules in pea is determined by the gene STIPULES REDUCED (ST), located on the linkage group (LG) III.The wild type has normally developed stipules (STST), while the double recessive (stst) genotypes are characterised by significantly reduced stipules (Fig. 2).There is the third allele, ST>BS, recessive to ST and dominant to st and responsible for genotypes with 'butterfly stipules', intermediate in size to normal and reduced stipules (Apisitwanich & Swiecicki 1992).
According to an early model of the pea leaf, the structure is determined sequentially during the repeated growth and sub-division of meristems, and secondly, the developmental fate of each meristematic primordium is determined by its size, that is, small primordia become tendrils, intermediate primordial leaflets and large ones rachides (Young 1983).The development of leaflets and tendrils in pea is determined by two genes, AFILA (AF) on LG I and TENDRIL-LESS (TL) on LG V.There are four main phenotypes determined by AF and TL genes (Fig. 3): (1) wild type (AFAF TLTL), with two or three pairs of leaflets and numerous tendrils; (2) 'afila' (afaf TLTL), with tendrils only; (3) 'clavicula' (AFAF tltl), with leaflets only; (4) 'pleiofila' (afaf tltl), with rachids ending with small leaflets.
The AF gene was described as a mutation by Kujala in 1953, while the TL gene was discovered by Vilmorin and Bateson in 1911.The gene AF influences pinna length and branching and regulates the timing and direction of leaf developmental processes in its proximal region (Villani & DeMason 1999).The TL gene is responsible for transforming lateral organ primordia from their default leaflet fate to a tendril fate where vascular bundles surround the central pith and abaxial-adaxial polarity is suppressed (Hofer et al. 2009).The wild-type TL allele probably arose as a semi-dominant mutation, surviving the constraints of selection by providing a novel phenotype and permitting Figure 2. Normal (left) and reduced (right) stipules in pea adaptation to a new and advantageous climbing growth habit (Hofer et al. 2009).There are reports on novel alleles at both AF (Vassileva 1979) and TL loci, namely TL2, only in the heterozygous state it transforms tendrils into very narrow leaflets (Berdnikov & Gorel 2005), and TL UL , with very narrow and elongated, entangling leaflets (Ukrainets 2009).

Agronomy of 'Semi-Leafless' Pea Leaf Type
In the early 1970s, it was suggested that the additional tendrils might improve the standing ability of the pea crop, leading to a wide use of the AFILA mutation (Snoad 1974).Its first application was the development of the fully 'leafless' pea (stst afaf) ideotype within a pea breeding program at the John Innes Institute and the release of the first UK 'leafless' cultivar Filby (JI 1768) in 1978 (Ambrose 2004).However, although the 'leafless' cultivars were able to produce sufficient photosynthate at normal agronomic densities (Hedley & Ambrose 1979), they failed to do so when grown at low planting densities through a limitation to the total biomass of plants and the crop itself (Hedley & Ambrose 1981).For this reason, the 'leafless' ideotype was largely abandoned and is mostly absent in contemporary pea breeding programmes.Introduction of the AF mutation with retained wild type stipules led to the development of 'semi-leafless' pea cultivars that proved superior to 'leafless' in photosynthetic capacity, similar to that of the wild type.This is considered perhaps the greatest achievement in pea breeding (Duparque 1996).The significantly increased standing ability of 'afila' pea cultivars reduced grain yield losses and associated reduction in canopy disease severity (Banniza et al., 2005) increased the interest in cultivating pea as a quality food and feed in many European countries (Mikić et al. 2006b) and in the world.Today, 'semi-leafless' pea cultivars make more than 95% of the total dry pea production in western Canada, about 80% in European Union and more than 30% in Russia.It has to be remarked that the 'afila' leaf type does improve lodging resistance, but on its own does not provide lodging resistance, where stiff stems, in addition to 'afila' leaf type, are also required.
Despite an initial impression that the 'afila' pea stand may be generally less productive, the opposite has been demonstrated in numerous occasions.A detailed physiological study of an 'afila' pea canopy (Kof et al. 2004) revealed that, on one hand, it had a decreased leaf area and a chlorophyll content per plant 1.5-fold lower than the wild type, while, on the other hand, the loss of leaflets was partly recompensed by expansion of the tendrils and stipules, extra accumulation of chlorophyll and increased assimilation area, resulting in enhanced photo-assimilating potential and non-diminished final biomass and grain yield.
The grain yield components of the 'semileafless' dry pea cultivars, developed for enhanced and stable dry grain production, do not differ significantly from those in pea cultivars with conventional leaves (Tab.1), especially if belonging to earlier cultivar generations with indeterminate stem growth (DET) and long internodes (LE) (Tab.2).This confirms that 'semi-leafless' pea cultivars, despite a somewhat slower initial growth and development, are at least equally productive as those with conventional leaves, with a considerable advantage regarding standing ability, increased photosynthetic activity of the whole canopy and less favourable conditions for pests and diseases.
An additional benefit of the 'semi-leafless' pea is development of autumn-sown, winterhardy cultivars, suitable for cultivation in diverse climates (McPhee & Muehlbauer 2007, McPhee et al. 2007), with emphasis upon warmer temperate regions.Such cultivars may contribute to a significant increase of the total pea cultivation area in many countries (Mikić et al. 2007).Average grain yields in autumn-sown 'semi-leafless' pea cultivars are at the same level as those in the spring-sown 'semi-leafless' and the autumn-sown with conventional leaves (Tab.3).Most autumn-sown 'semi-leafless' pea cultivars are also characterised by a prominent earliness in comparison to the spring-sown cultivars (Fig. 4).Thus they offer more economic benefit to farmers, especially in central or southern European countries, who often have only one combine harvester available and have to choose between pea and cereals, as a rule giving priority to the latter.Now, with autumn-sown, early 'semi-leafless' pea cultivars, they are provided with a possibility to harvest pea first and then barley or wheat (Mikić et al. 2011).Pea leaves with large surface areas were not a priori large in terms of their dry mass (Niklas et al. 2008).This opened the possibility to develop 'semi-leafless' cultivars for high and stable forage production (Koivisto 2003, Warkentin et al. 2009), as well as to solve a long-term problem of reliable seed production in forage pea cultivars, since those with conventional leaves often suffer from excessive lodging and high seed losses during the harvest (Karagić et al. 2009).Modern 'semi-leafless' forage pea cultivars (Fig. 5) are characterised by a plant height up to 100 cm, thick and juicy stems, large number of internodes, large stipules comprising up to 76% of the total leaf area (Goldman & Gritton 1992b), tendrils contributing up to 13% of total dry matter (Uzun et al. 2005) and pods grouped in the upper half of a plant (Mihailović et al. 2009).Preliminary research (Ćupina et al. 2010) has demonstrated intercropping pea cultivars with conventional and afila leaves may prove economically justified (Tab.4), offering an alternative to the traditional intecrops of pea with oat (Avena sativa L.) or barley (Hordeum vulgare L.).

Agronomy of 'Acacia' and Other Pea Leaf Types
The agronomic value of the 'acacia' pea cultivars with 'clavicula' leaf type is generally much less studied in comparison to the 'semileafless' cultivars or those with conventional leaves.The main reason for this long-term ignorance by agronomists may lie in the fact that such genotypes are extremely prone to lodging due to the lack of tendrils and therefore are immediately recognised as risky for obtaining high and stable grain yields, although their main grain yield components do not have to differ significantly from those in the cultivars with conventional leaves (Tab.5).The 'acacia' pea genotypes could be interesting in breeding for forage yield (Mihailović & Mikić 2004), with an ideotype with large number of nodes, large stipules, large number of photosynthetically active leaves and pods grouped in the upper half of the plant in order to minimize the losses during harvest (Fig. 6).However, it is still not clear if increasing the total leaf area will lead to a higher forage dry matter production and if an increased leaf proportion in the total forage yield will also result in a higher forage dry matter crude protein content and thus better forage quality.These questions need to be solved by breeding programmes aimed at the development of 'acacia' forage pea cultivars.At this moment, it is likely that such genotypes could be safely cultivated for forage in mixtures with cereals that would serve to support them, since the lodging of 'acacia' pea plants is significant at the stages of full bloom when the crop is usually cut for forage.
Pea with the 'pleiofila' leaf type is a fine example of two-gene interactions where both equally contribute (Marx 1987).In this case, the af allele affects marginal meristem function in the leaflet, while the tl allele gives rise to adaxial and marginal meristems where none existed before (Meicenheimer et al. 1983).Leaves on the 'pleiofila' plants increase in complexity more rapidly during shoot ontogeny than those on plants with conventional leaves, while leaflets of 'pleiofila' plants have identical histology to wildtype leaflets although with smaller and fewer cells (Villani & DeMason 1997).
Reports on the agronomic performance of the 'pleiofila' pea genotypes are scarce and insufficient for a more detailed comprehensive understanding of their breeding and agronomic value.When grown at a standard highly competitive population density, they do not show any yield improvement compared to those with conventional leaves (Goldman et al. 1992a), however, when grown under minimum competition and in the latest maturing background, the 'pleiofila' pea lines may exhibit the largest increases in grain yield in comparison to the genotypes with conventional and 'afila' leaves (Goldman & Gritton 1992, Goldman et al. 1992b).
A few additional leaf mutants in pea could be economically interesting.Among the most interesting examples is a novel mutation at the AF locus, weaker than existing alleles, resulting in a pair of leaflets of normal size and morphology or a single leaflet at or just after the ramification of the central rachis (Fig. 7).Such genotypes, referred to as 'semi-leafless+', may be interesting for breeders as the presence of a pair of leaflets at each node may increase the photosynthetic capacity of the plant, enable further increases in biomass at lower densities and increase the leaf area index of a crop early enough in crop development to cut light levels penetrating the crop and thus help suppress emerging weeds (Ambrose 2004).Another mutant, developed at the Institute of Forage Crops in Pleven has four pairs of leaflets on every node and could be interesting for forage production, although the stability of this trait needs to be assessed in diverse environments (Ellis et al. 2009).

Conclusions
Pea has a unique place among the economically important legume species due to its variable leaf morphology.Its genetic background is well studied and provides breeding and other applied research with diverse beneficial possibilities.It may be claimed that the 'semi-leafless' pea cultivars demonstrated the highest level of agronomic utility, playing an essential role in the maintenance and improvement of dry pea production across the world.Other leaf types in pea are less studied but deserve additional evaluation by both geneticists and agronomists, especially in the case of 'acacia' that could have a potential for forage production and 'semi-leafless+' for higher productivity.

Figure 1 .
Figure 1.The uni null mutant in pea with normal stipules and only one leaflet (Pisum Genetics Association 2011)

Figure 4 .
Figure 4.The latest Serbian winter dry pea line L-574 (right), registered in November 2010 under the name of the cultivar Mraz, in comparison to the widely used Serbian spring-sown dry pea cultivar Jezero (left), photographed on the same day, May 12, 2009(Mikić et al. 2011)

Figure 7 .
Figure 7.A detail of the leaf (left) and the stand (right) of the 'semi-leafless+' line 11/47 AFILA (JI 3129) developed at the John Innes Centre

Table 5 .
Average values of grain yield components and grain yield of pea cultivars with different leaf types at Rimski Šančevi during 2000-2002(Mihailović & Mikić 2004) Figure 6.A typical 'acacia' pea plant