It seems I have a fascination with legumes. I was about to pass up the New Phytologist journal’s booth when they handed me a reprint of the recent (2017) review by Sprent, Ardley, and James on legumes. Before I read this article, if you had asked me what a typical legume is I probably would have answered “beans”. However, as the third largest plant family, the legumes present a kaleidescope of forms, most of which are not your garden variety.
The legumes are defined by pods that open up along both edges, labeled by botanists as “loments” or “legumes”. Many legumes host nitrogen-fixing bacteria, but many do not, and not all plants with nitrogen-fixing partners are legumes. In my classes I learned that there were three subfamilies — or maybe four — that were classified according to their flower type. One subfamily was characterized by flowers that are pattered like those of the pea, with a large petal at the top giving the appearance of a bonnet, the other petals forming something like a beak in front of it. An early botanist had the impression of a butterfly when he saw this type of flower and called it papilionoid, and thus the subfamily with this type of flower became the Papilionoideae. A second subfamily had flowers with five more-or-less equal petals, sometimes with “claws”, and with a slight bilateral symmetry, such as those my grad school advisor had studied in the tropics. Her connection to these species left me with certain regard for this group, the Caesalpinioideae, named in honor of another early botanist. The third subfamily had flowers with no petals, only stamens and pistils, that clustered together into fuzzy puffs such as those on an acacia, a group named the Mimosoideae, evoking a certain brunch cocktail. The fourth group designated as a legume subfamily had been recognized as a separate family by the time I learned of it.
Part of the reason that legumes are so successful, and the reason they are important to humans, is their ability to harbor specialized bacteria that turn nitrogen gas into nitrogen fertilizer, a process known as nitrogen fixing. The bacteria benefit from this arrangement because they are fed and protected inside the legume roots, while the legume gets millions of tiny internal fertilizer factories. In another one of my classes I was taught how the plant sends a chemical signal into the soil that attracts the bacteria. The bacteria respond with their own chemical signal, causing the plant root hair to curl around a single bacterium and open up a narrow tube, called an infection thread, for the bacterium to enter. The bacterium travels into the main part of the root, where it multiplies as the plant builds a nodule around it. The plant releases the bacteria into a specialized compartment, where they become nitrogen-fixing zombies called bacteroids that have lost the ability to divide, often becoming unable to live in the soil again once the root dies and the nodule breaks open.
Sprent, Ardley, and James have updated my understanding of legumes and nodulating bacteria. The agricultural legumes we are familiar with, especially those grown in temperate zones, are biased toward European and Middle Eastern species such as alfalfa, clover, vetch, and peas, which have been exported around the world. In other parts of the world, by contrast, legumes are often trees or shrubs. The textbook infection-thread story is just one mode of nodulation. In some legume species the infection thread does not end up anywhere, becoming a “fixation thread” as the bacteroids start performing their nitrogen fixation inside the tube. In other legume species the bacteria enter through the root epidermis, and no infection thread is formed. This is the mode in the lineage that includes Scotch broom, rooibos, sunnhemp, and lupine, the latter being noteworthy for its wraparound nodules. Finally, some legume species allow bacteria to enter through cracks. A semi-aquatic West African Sesbania species grows nodules in this way — on its stem.
The previous classification scheme has been overturned, as evidence from DNA has shown that the Mimosoideae are actually a branch of the Caesalpinioideae. Furthermore, some members of the Papilionoideae have perfectly radially symmetric petals, giving the impression of a hibiscus flower. All of the African “acacias” are now placed in separate genera (sing. genus) from the Australian genus Acacia. And even though I had an image of caesalpinioids as tropical trees, the genus Chamaecrista includes temperate zone species and annuals, such as partridge pea, a North American native used for forage.
Another principle I learned from classes was that a species of nodulating bacteria is specific to a legume host genus. The specificity may be very strict — soybean will not nodulate in North America unless the soil is inoculated with the bacterium Bradyrhizobium japonicum. This trend does indeed hold widely across the legumes, particularly those widely exported European/Middle Eastern species. However, among Mimosa species, it is soil factors that determine the symbiotic bacterial species. Sometimes when a legume plant genus is distributed between different continents, its various member species adapt to whichever class of nodulating bacteria is present on the respective continent. There are also legume species that are “promiscuous”, such as the common bean, that can nodulate with bacteria belonging to the distantly related classes.
One mechanism for host-bacteria specificity seems to be a result of evolutionary arms race dynamics. The species of legumes in the European/Middle Eastern group produce a suite of antimicrobial-like molecules that regulate the process of turning free-living bacteria into highly efficient and subservient bacteroids. Each bacterial species, for its part, produces a unique enzyme that neutralizes a range of these regulatory molecules. On the one hand, the enzyme can allow the bacterium to multiply more successfully in the plant and get away with fixing less nitrogen. On the other hand, the particular set of regulatory molecules that the enzyme neutralizes will determine which hosts it can exist within.
Another factoid from the classroom is that some legumes are capable of producing potent toxins. A lecturer told of an acquaintance who cooked and ate scarlet runner bean roots and had to have his stomach pumped. Someone else became ill from munching on a single raw scarlet runner bean while working in her garden. Raw kidney beans can cause nausea with the ingestions of single-digit quantities. Consider, though, the southwest Australian leguminous shrub genus Gastrolobium, which produces fluoroacetate, a cell respiration inhibitor so toxic as to eliminate a sheep after just a few bites. Evidently the endemic legumes of this area share an unusual class of toxins that are not alkaloids and do not incorporate nitrogen atoms at all. Perhaps this attribute is related to their adaptation to low-phosphate soils, phosphate being crucial in most legumes for the nitrogen fixation process.
Some lineages of nodulating bacteria have given rise to well known pathogens. Agrobacterium, associated with crown gall, was revealed to be part of the genus Rhizobium. Agrobacterium uses a small circle of DNA separate from the main chromosome to carry its infection genes, just as nodulating Rhizobium species use a separate DNA circle for their nodulating genes. Another genus, Burkholderia, is known for several pathogenic species causing blights and spots on crops and ornamentals, and only fairly recently was discovered to include nodulating species found in South American mimosoids and South African papilionoids.
The different classes and genera of nodulating bacteria are often associated with particular environments. For instance, Burkholderia species prefer acidic soils and higher altitudes, whereas in higher-nutrient soils they are outcompeted by other species of nodulating bacteria. Other genera are associated with seasonally dry acidic soils or with alkaline soils. The genus Cupriavidus is known for heavy metal tolerance. However, strains of Rhizobium can be found in many environments.
A couple more factoids: Although the soybean, Glycine max, is native to China, oddly the genus Glycine is most diverse in Australia. And aside from providing nitrogen, nodulation may be beneficial in other ways, such as improved water use efficiency. I wonder how this benefit would figure into Ford Denison’s analysis of “cheater” strains. Myriad other factoids are presented in the review that only a legume researcher could love. If you know what “mirbelioid” means, you are the target audience. The value of knowing the full diversity of the legumes, according to the authors, is to help in the search for legumes adapted to extreme conditions for use in agriculture under a changing climate. I would also add that new agricultural legumes would be useful if they have resistance to diseases of today’s set of forage and green manure legumes.
So what is a typical legume? The legume genus Astragalus, made up of herbs and small shrubs, the milkvetches and locoweeds, contains something like 2,500 species, the most of any plant genus, but it is hardly representative of the legume family. I would have to say that this is another instance of a short question with a long answer.
Since you asked… We’ve defined a rhizobial “cheater” as a strain that benefits itself by providing less benefit to the host than the strains that would otherwise occupy its nodule (Kiers & Denison 2008). So it depends on *how* rhizobia increase water-use efficiency. The obvious way is to provide nitrogen that increases leaf photosynthesis relative to transpiration (p.128 in my book). Assuming that mechanism, a cheater could be a strain that diverts more resources from nitrogen fixation to its own reproduction than other local strains, or one that manipulates the host into giving it more resources than other strains with similar nitrogen fixation, or makes the host make more or fewer nodules than it needs in a way that increases its own nodulation opportunities relative to competing strains. A more interesting mechanism would be to manipulate the plant into closing its stomata in the afternoon, when water-use efficiency is low. This would benefit the plant community as a whole (p.130 in my book) and so benefit local rhizobia collectively, but I don’t see how the individual manipulator would benefit. So I’d call it an altruist (unlikely to persist) rather than a cheater.
I remember how traditional Mayan farmers in southern Mexico had ways of recognizing the different ways legumes operated in their farming systems, covering all of the types you describe above. They didn’t know the actual biological mechanisms, but they had observed their impacts over generations. It is fun to think about putting the two types of knowledge together.