Why is a raven like a writing desk? Origins of the sunflower that is neither an artichoke nor from Jerusalem


The Mad Hatter asks a nonsensical riddle at his tea party where Alice drops in on her way through Wonderland – Lewis Carroll later contended that the conundrum had no answer, although many have tried to find one. Polyploid plant complexes, like the one to which the Jerusalem Artichoke belongs, have long been seemingly impenetrable riddles like that of the Mad Hatter, but Bock et al., in this issue of New Phytologist (pp. 1021–1030) use a new molecular technique coupled with deep understanding of the plants themselves to unravel this complex story. The origins of this ‘minor’ cultivated plant have been elucidated, and very elegantly too. The Jerusalem Artichoke, Helianthus tuberosus, a member of the daisy family and of the same genus as the cultivated sunflower Helianthus annuus, illustrates well the vagaries of common names – once coined they stick. Investigation of the origins of cultivated plants can be fraught with problems, among them common names that can mislead.

‘The organellar genomic fraction in Helianthus revealed NO species were reciprocally monophyletic  – something that would make many biologists shudder – the tree is a mess!’

The Jerusalem Artichoke, despite its name, is one of only a few plants domesticated in North America, a region not otherwise a hotbed for crop origins (Meyer et al., 2012). Although only recently making a comeback on supermarket shelves, the Jerusalem Artichoke was very popular in Europe before the potato (Solanum tuberosum, brought from Peru by the Spanish) caught on in the mid-fifteenth century. Its resurgence in popularity has to do with its purported health benefits and its climatic versatility (see references in Bock et al.). As a species, human beings depend upon fewer than a dozen plant species for 80% of their calories and in the light of concerns over global food security and a growing human population (McCouch et al., 2013), interest in both a range of new crops and in the wild relatives of crops is increasing. One perennial problem with many domesticated plants is that the relationships with their closest relatives are still sometimes obscure, despite years of research and active plant breeding. This is often due to the complexity of their origins and to the active spread of crops worldwide post-domestication with human spread around the globe.

Many of our major crops are polyploids, with heritable increases in chromosome numbers relative to other species related to them. Polyploidy is extremely common in plants, it has been estimated that 15% of all flowering plant speciation events are accompanied by ploidy increase (Wood et al., 2009) and the diversification of major angiosperm lineages has been attributed to major genome duplications (Tomato Genome Consortium (TGC), 2012). Many of our most important crops are polyploid – wheat, potatoes and cotton to name a few. Because these major crops are polyploid, it has been thought that ploidy changes were a characteristic of plants under domestication but recent analyses of a broad sample of domesticated plants (Meyer et al., 2012) show that the proportion of polyploid crops (17%) is comparable to that in angiosperms in general. Polyploidy confers instant reproductive isolation and profound later effects due to genome duplication and subsequent divergence and gene neofunctionalization (TGC, 2012). Polyploidy comes in two basic flavours in plants: autoploidy, which is a simple doubling of chromosome number in a species creating two chromosome complements that are identical; and alloploidy, where the genomes of two (or more) different species come together to form a new chromosome set, a genome from both parents. It can also get more complicated than this, as Bock et al. show.

One recurring difficulty with phylogenetic analysis of groups containing polyploid species is the fact that allopolyploid species are the result of hybridization, not monophyletic, and thus cause reticulation on a strictly dichotomously branching tree. Early in the development of phylogenetic systematics it was suggested that all lineages containing these ‘problem children’ be collapsed to polytomies (Humphries, 1983), but given the widespread occurrence of polyploidy in plants, more recent studies have focused on the origins of the taxa, not the way in which the tree is drawn (Knapp et al., 2004). The Jerusalem Artichoke presents a particularly intriguing case where the polyploid domesticate itself is a member of a lineage that contains several wild polyploid relatives, and where origins of the cultivated species have long been debated.

Standard sequencing can be used to elucidate the origins of polyploids by combining regions from maternally inherited plastid components and paternally or biparentally inherited nuclear components (Chase et al., 2003). This falls apart though in groups where recent, rapid radiation has occurred, and where hybridization between species is common at all levels. These factors have made the untangling of Jerusalem Artichoke origins using standard sequencing techniques inconclusive (Timme et al., 2007). Bock et al. break through this barrier by using the new technique of genome skimming to crack the riddle. Genome skimming (or my less preferred synonym, ultra-barcoding, e.g. Kane & Cronk, 2008) is a technique for assembling and analysing the high-copy genomic fraction of plastid, mitochondrial and nuclear ribosomal DNA (rDNA) from next-generation sequencing technologies to provide large amounts of data from these complementary marker categories. This is really a case of more data making the difference, rather than more taxa.

Helianthus is a genus with a number of polyploid perennials in North America as well as the annual cultivated sunflower H. annuus. The crossability of the sunflower to H. tuberosus (Jerusalem Artichoke) suggested it might be the wild progenitor, but Bock et al. analysed all possible scenarios and it is clear that H. annuus is not part of the history of the Jerusalem Artichoke. This highlights an important reason why botanists have never really taken to the biological species concept (BSC) as articulated by Ernst Mayr (1992) and taught in all introductory biology classes (Knapp, 2008); plants seemingly cross willy-nilly. But the ability to cross is a plesiomorphic (‘primitive’ or ‘ancestral’) trait and thus of limited utility in elucidating relationships. Ability to cross is important for plant breeding however, and so much of the literature on crop wild relatives assesses relationships with this criterion (Harlan & de Wet, 1971). A crop wild relative (sensu Maxted et al., 2006) might not always be evolutionarily most closely related to a crop! New methods, like the genome skimming one used by Bock et al., will help us to separate these different aims and focus more clearly on the biological processes involved. Better evolutionary understanding will help with the prioritization (Vincent et al., 2013) and subsequent use of crop wild relatives in plant breeding in the face of widespread environmental change.

But in the end, all the molecular sequencing in the world will not solve complex problems like the origins of the Jerusalem Artichoke in the absence of knowledge of the organism. Seemingly about a new technique applied to an unanswerable riddle, Bock et al. really show that understanding the plants themselves is critical for solving biological problems. The organellar genomic fraction in Helianthus revealed NO species were reciprocally monophyletic – something that would make many biologists shudder – the tree is a mess! Bock et al. suggest that this is the result of incomplete lineage sorting – the retention of ancestral polymorphism due to rapid radiation of Helianthus. Here is where knowledge of the organism comes in; it makes sense in the light of the biology of sunflowers. Other polyploid lineages will have other problems involved with their own biology (Kovarik et al., 2004) there is no one-size-fits-all recipe, however much we would like there to be.

With the advent and uptake of new sequencing techniques where ever larger portions of the genome are available for analysis it is becoming clear that the genome itself is a complex ecosystem, full of signals of the organism's evolutionary past. What is exciting is that we are now developing the tools to explore this new unknown and that the sorts of studies undertaken by Bock et al. raise as many questions as answers – a hallmark of the best science to my mind. For example, why did the Jerusalem Artichoke form multiple times? What was good about that combination of genomes? … My sense of wonder at the sheer complexity and elegance of life on Earth is constantly reinforced by studies such as this – life is indeed a ‘tangled bank’ as Charles Darwin (1859) so aptly put it. As we delve deeper into the genomic organization of nonmodel species and deploy this information in ever-increasingly sophisticated ways in a robust evolutionary and phylogenetic framework we come closer to untangling the bank and providing plausible and robust evidence to answer what were previously intractable and unanswerable riddles.