Research priorities and opportunities for forage and feed crops in Mediterranean countries

Crop genetic improvement is expected to play a pivotal role in strategies of climate change adaptation and mitigation (Olesen and Bindi, 2002; Ceccarelli et al., 2010), also in view of the higher sustainability of rising yields through improved germplasm than through more expensive and/or possibly environment-unfriendly inputs (e.g. mineral fertilizers; pesticides; irrigation). The limited funding available at the national and international level for forage and feed legume breeding requires to carefully identify a few promising species on which concentrating joint efforts and resources. Ideally, the target species should feature: i) wide adaptation to soil, climatic and management conditions of the Mediterranean basin; ii) wide flexibility of utilizations; iii) high drought tolerance and water use efficiency (WUE), intrinsically and by growing extensively in the wet season; iv) at least some knowledge by farmers. In addition, regional research should avoid duplication of work carried out by the International Center for Agricultural Research in the Dry Areas (ICARDA), whose breeding activity has targeted grass pea (Lathyrus sativus), some vetches and faba bean (the latter mainly as a food legume). On the grounds of these considerations, this project has identified lucerne (alias alfalfa; Medicago sativa) among the perennials, and pea (Pisum sativum) among the annuals, as the target legume species for breeding work and research, contemplating also research for optimization of their use in legume-based forage mixtures.

Lucerne is the main forage crop in south-European countries such as Italy and France, as well as in Morocco. Its cultivation in Maghrebian countries is traditionally limited to oasis and frequently irrigated conditions (, but results of the EU-funded project PERMED ('Improvement of perennial forage plants for sustainability of Mediterranean farming systems' – INCO PL509140) have highlighted the good adaptation to severely drought-prone environments of Mediterranean landraces that evolved in stressful environments (Annicchiarico et al., 2011a; Annicchiarico et al., 2013). The outcome of this work has allowed to establish a genetic base at CRA-FLC of Lodi, which can be used as a mapping population and for phenotypic selection targeted to drought-prone regions.

Lucerne is traditionally grown under a mowing regime in monoculture, to produce high-protein hay, silage or fresh fodder. However, it can also be grown in mixed stand with a forage grass under a mowing regime, to improve the crop yield and seasonal distribution and the efficiency of nitrogen utilization while reducing the nitrate leaching and the weed invasion (Huyghe and Tabel, 2010). PERMED provided useful results also as regards candidate drought-tolerant grass varieties for mixed cropping with lucerne (Annicchiarico et al., 2011b; Pecetti et al., 2011). In addition, lucerne varieties can be developed specifically for grazing systems (Pecetti et al., 2006). Breeding such varieties for Mediterranean regions may require, however, to break the inverse genetic correlation of grazing tolerance with low autumn dormancy (Pecetti et al., 2006).

The on-going development of lucerne genomic resources (genome sequences and SNP markers) at the Samuel Roberts Noble Foundation (SRNF) and at INRA-URP3F of Lusignan may allow to explore association mapping also for quantitative trait loci (QTL) (Li et al., 2011) and, possibly, even genomic selection procedures (Heffner et al., 2009). A Genotyping-by-Sequencing approach (Elshire et al., 2011) which is under development for lucerne at SRNF may soon enable to analyze sequence variation in several thousand loci across the genome in many individuals of a population. Also candidate gene approaches have proved to be efficient in lucerne (Herrmann et al., 2011). QTL for lucerne drought tolerance have just started to be studied within the project PERMED (Julier et al., 2010). Marker-assisted selection (MAS) procedures would be valuable also to breed grazing-tolerant lucerne with low autumn dormancy, or to select salt-tolerant varieties.

Pea is the main feed grain legume in Europe, while being mainly grown for forage in mixture with a cereal in the Maghreb ( It has remarkable flexibility of utilization, as it may be harvested at crop maturity for grain (used as a concentrate) and straw (usable as a fodder), or harvested earlier, for hay or silage production. In addition, it may be grazed at maturity if unfavorable climatic conditions led to poor grain yield. Pea grains are valuable as concentrates for poultry, livestock or pigs because of their high protein and energy value, lack of antinutritional factors and ease of conservation (UNIP-ITCF, 2001).

A traditional drawback of pea, namely its poor standing ability, has been improved remarkably by recent plant breeding. Novel varieties, however, have hardly ever targeted regions of the Mediterranean basin. Nevertheless, pea has showed higher grain yield than faba bean or lupins in southern Europe (Annicchiarico, 2008), as well as better adaptation than faba bean, grass pea or various vetch species to semi-arid Australian environments (Siddique et al., 2001). The evaluation of a large number of recent pea cultivars in Mediterranean environments of Italy (Annicchiarico and Iannucci, 2008) has allowed to identify elite parent material which has been used to produce sets of recombinant inbred lines (RIL). Genomic resources (mainly SNP markers) are increasingly available for pea genotyping work aimed to define MAS procedures (Deulvot et al., 2010).

Bulk segregant analysis is a promising approach for identifying QTL associated with tolerance to one major stress which also implies the simultaneous selection of novel stress-tolerant germplasm. Its only documented application in legume tolerance to abiotic stresses has been provided by Castonguay et al. (2006) for lucerne cold tolerance.

More efficient breeding may arise not only from MAS procedures but also from ecological and/or evolutionary approaches that exploit selection under natural or artificially-reproduced stress conditions (Ceccarelli et al., 2010). One such approach has been applied successfully to lucerne breeding in Italy, allowing to breed specific varieties for moisture-favourable and drought-prone areas (Annicchiarico, 2007) after exploiting the ability of managed environments to reproduce the genotype responses to specific agricultural environments (Annicchiarico and Piano, 2005).