Parasitic plants, wild relatives and the nature of resistance

Many flowering plant species have lost their autotrophic way of life during evolution – they parasitise other plants in order to supply themselves with water and nutrients. Several are severe constraints to the growth of major crops, and pragmatic methods of control have been used that centre on agronomic practices and the use of herbicides. However, success has been limited. Now, the focus has shifted to take account of recent developments in crop breeding and molecular genetics, and the need to develop genetically resistant cultivars. Not only is this proving immensely successful agronomically, but it is also providing insights into the basic mechanisms of plant parasitism and resistance. In this issue of New Phytologist (pp. 557–568), Gurney et al. describe resistance to the parasitic plant Striga hermonthica in Tripsacum dactyloides, a wild relative of maize. Approximately 3500 flowering plant species have become parasites, becoming nonautotrophic and gaining their nutrition through special multicellular organs, the haustoria, from other plants. By far the most economically damaging parasitic weeds are members of the genera Striga (witchweeds) and Orobanche (broomrapes). Striga spp. are a major pest throughout semiarid Africa and parts of Asia. S. hermonthica and S. asiatica infect important cereals such as maize, sorghum, millet and upland rice, threatening grain production for subsistence farmers in Africa. S. gesneroides is a serious constraint to cowpea production in Africa. Orobanche spp. are important in southern and eastern Europe, the Middle East and North Africa: O. crenata causes huge damage to legume crops. O. cumana threatens sunflower. O. ramosa attacks potato, tobacco, tomato and hemp. O. aegyptiaca has very severe effects on legumes and vegetables. The main focus of research on parasitic weeds has been on their management when infecting important crops. Control strategies have centred around agronomic practices and the use of herbicides, although success has been marginal. There is therefore an urgent need to re-evaluate control methods in the light of recent developments in crop breeding and molecular genetics and to place these within a framework that is compatible with current agronomic practices. There is a strong consensus that growing genetically resistant cultivars is the most appropriate and cost-effective means of managing pest and diseases (parasitic weeds among them), and it is one of the key components of crop improvement. Where has this consensus led so far? Development of improved cultivars with resistance to a single pathogen is often straight-forward if a good source of resistance is available and an efficient, easily controlled and practical screening procedure exists to provide good selection pressure. Unfortunately, this is seldom the case with parasitic weeds. Resistance against most parasitic weeds is difficult to access, scarce, of complex nature and of low heritability, making breeding for resistance a difficult task. In spite of these difficulties, significant success has been achieved in some crops. In a few instances, resistance of simple inheritance acting after parasite penetration has been identified and widely exploited in breeding. This has been particularly important, allowing a rapid progress in sunflower breeding against O. cumana and cowpea breeding against S. gesnerioides. However, breeding programmes based on only a few dominant genes are at serious risk from breakdown of resistance. A well studied case is O. cumana attacking sunflower, in which up to seven races have been described. This breakdown of formerly effective resistance, although not as rapid as experienced with other pathogens, is the nightmare of breeders, and forces them to search continuously for new sources of resistance (Fernández-Martínez et al., 2000). A similar situation might be happening in S. gesnerioides attacking cowpea, with several races reported (Lane et al., 1997). Only moderate to low levels of incomplete resistance of complex inheritance against parasitic weeds have been identified in other crops, such as in maize against S. hermonthica or legumes against O. crenata. This has made selection more difficult and has slowed down the breeding process, but the quantitative resistance resulting from tedious selection procedures has resulted in the release of cultivars with useful levels of incomplete resistance combined with a degree of tolerance (Cubero et al., 1994; Berner et al., 1995; Haussmann et al., 2000). This resulting resistance might be a combination of mechanisms of resistance more likely to last longer than resistance based on single genes. In fact, although some variation among populations exists, no clear evidence for existence of races has been found in these parasitic weeds. The key to the development of cultivars with long-lasting resistance might be diversity. This might include diversity in the type of resistance, and diversity in their strategy of deployment. A wider genetic base to the resistance is essential to ensure durability. Resistance is a multicomponent event, being the result of a battery of escape factors or resistance mechanisms acting at different levels of the infection process. Resistance commonly exploited in breeding acts after haustorium formation, either by encapsulation or necrosis of the tubercle or of the host tissue. There are similarities of this process with hypersensitive cell death found in many pathosystems with a role of phytoalexins and pathogenesis-related proteins (Jorrín et al., 1999; Joel, 2000). Other mechanisms of resistance have commonly been neglected. An exception is the exploitation of low induction of germination of Striga seeds in sorghum breeding (Haussmann et al., 2000). This mechanism was, however, not found in other crops such as in cowpea, maize or tropical grasses (Lane et al., 1997; Kuiper et al., 1998; Gurney et al., pp. 557–568 in this issue), but has recently been identified in legumes against O. crenata (Rubiales et al., 2003) and in Helianthus against O. cumana (Jorrín et al., 1999; Labrousse et al., 2001). Host plants might escape infection by reduced root biomass and by root architecture that avoids the soil layer in which the seeds of the parasite are more common. The host may limit damage (tolerance) by factors that influence source–sink relationships, such as osmotic pressure. There is variation for these traits in germplasm collections. Landraces have been widely replaced by homogeneous cultivars that may not possess the same combined tolerance and resistance levels (also present in wild accessions). Genetic resources remain highly unexplored and underused and might contain very valuable additional sources of resistance. A battery of resistance mechanisms against O. cumana has been recently described in wild species of Helianthus (Labrousse et al., 2001) and is being transferred into cultivated sunflower. Similar approaches for identification and exploitation of genetic resources in breeding against parasitic weeds have been followed in other crops (Kuiper et al., 1998), and this includes the work of Gurney et al. Gurney et al. show that Tripsacum dactyloides, a wild relative of maize, is not only resistant to Striga, but also that partial resistance is inherited in a hybrid between T. dactyloides and maize. Dissecting the escape and resistance factors will help to detect existing genetic diversity for mechanisms that hamper infection (Jorrín et al., 1999; Haussmann et al., 2000; Labrousse et al., 2001; Rubiales et al., 2003). Combining different resistance mechanisms into a single cultivar will provide a durable outcome. This can be facilitated by the use of in vitro screening methods that allow the dissection of parasitic weed resistance into highly heritable components, together with adoption of marker-assisted selection techniques (Haussmann et al., 2000; Román et al., 2002). This presents intriguing agronomic possibilities.