Cooperation gets complex
This week, instead of discussing a paper, I will summarize some presentations from the Ecological Society of America meetings last week in San Jose, California (a much more interesting place than I expected, including hands-on transformation of bacteria at the Tech Museum).
I ran into two people who admitted to reading this webblog: Madhu Katti and Don Strong, but didn't get to talk to either of them for long. There were usually several interesting sessions going on at once, from 8 AM to late evening, plus lots of informal discussions, but I will limit my summary to a few of the presentations on the evolution of cooperation, my own area of research.
Rhizobium bacteria and mycorrhizal fungi both infect plants. Both are often beneficial, providing host plants with nitrogen or phosphorus, respectively, "in exchange for" carbon. But is this really an exchange, in the sense that there is some mechanism to ensure that a rhizobium cell gets a certain amount of C from the plant if and only if it provides a certain amount of nitrogen to the plant? This seemed to be an implicit assumption in the "market model" of Claire de Manzancourt and Mark Schwartz, presented at the meeting, but this hasn't yet been demonstrated even in rhizobia, much less in other species.
Providing the plant with nitrogen increases plant growth, which may automatically increase carbon supply to the rhizobia, collectively. But if these benefits were shared indiscriminately with every rhizobium strain infecting a given legume plant -- strains which will soon be competing for the next host plant! -- then modeling by Stuart West on a visit to my lab showed that natural selection would favor little or no investment by rhizobia in nitrogen fixation.
Our model assumed that carbon that is not respired by a rhizobium cell to power nitrogen fixation can instead be used to support the rhizobium's current or future reproduction. My students and I addressed various aspects of this assumption at the meeting.
Other researchers (Cevallos et al. 1996 and Hahn and Studer 1986) have shown that there is a trade-off for rhizobia between fixing nitrogen and accumulating the energy-rich molecule, polyhydroxybutyrate (PHB). But does PHB help rhizobia reproduce? In his talk, Will Ratcliff showed that it does. He used a centrifuge to separate genetically identical rhizobia into low- and high-PHB fractions. Under starvation conditions, the high-PHB rhizobia reproduced two-fold, using up their PHB in the process, before starting to die off. The low-PHB rhizobia started dying immediately. So rhizobia that accumulate more PHB might be expected to out-compete those that fix more nitrogen. But read on!
Ryoko Oono's poster presentation discussed some legume plant species that prevent nitrogen-fixing rhizobium cells from reproducing inside their root nodules. Like worker bees, these rhizobia will have no direct descendants, so hoarding PHB would be pointless -- so they don't. But why do rhizobia give up the ability to reproduce? Once inside a root nodule, the plant may be in charge, but then they why do rhizobia infect plants in the first place? Because not all of the rhizobia inside the root nodule change into the nitrogen-fixing, nonreproductive form. Apparently, enough reproductive clone-mates eventually escape the nodule into the soil that a rhizobium cell founding a nodule will still have more descendants than if it stayed in the soil.
Plants that limit rhizobium reproduction typically have nodules with indeterminate growth, leading to an elongated shape. By reducing the incentive for rhizobia to accumulate PHB, at the expense of nitrogen fixation, suppression of rhizobium reproduction should benefit plants. But the ancestor of all nodulated legumes is thought to have had indeterminate nodules. If indeterminate nodules are so great, why would beans, birdsfoot trefoil, etc., switch to roughly spherical, determinate nodules, in which rhizobia retain the ability to reproduce? Ryoko looked at rhizobia in indeterminate nodules of species thought to resemble ancestral legumes and found no evidence that rhizobium reproduction is suppressed in these species. Apparently only a subset of legumes with indeterminate nodules suppresses rhizobium reproduction. Maybe in a few million years this plant trait will be more widespread, although a few rhizobia have apparently evolved a new trick to beat this system: rhizobia that are able to fix nitrogen transfer some carbon to their reproductive sisters instead.
What keeps this trick (or PHB hoarding, by reproductive rhizobia) from spreading? Toby Kiers, previously in my lab at UC Davis, showed that rhizobia that fail to fix any nitrogen are subject to host plant sanctions that greatly reduce their reproduction, but moderate cheating by rhizobia may be punished much less severely (Kiers et al. 2007).
Katie Heath (working with Peter Tiffin here in Minnesota) found no evidence of sanctions against any of three rhizobium strains from a plant with indeterminate nodules. The three strains differed in the benefits provided to a given host plant genotype, but some of that difference could have been due to limited ability to nodulate, which would not be expected to trigger sanctions, rather than differences in nitrogen fixation rate per nodule. Maren Friesen (UC Davis) commented that severe cheating (fixing no nitrogen) is punished in this species, as indicated by much smaller nodules containing few rhizobia.
Do sanctions also ensure fair trade between plants and mycorrhizal fungi? In his talk, Jim Bever (Indiana University) showed that a plant with each half its roots infected by a different fungus sends more carbon to the side with the fungus most beneficial to the plant. But when two fungal species are thoroughly mixed in the soil, the strain that is worse for the plant sometimes wins. He concluded that spatial structure is important to maintaining more beneficial mycorrhizal fungal species. The question is whether the degree of spatial structure in natural soils is more similar to the half-root experiment or the mixed-soil experiment. What about agricultural soils?
Speaking of agriculture, my favorite talk was by Cameron Currie (University of Wisconsin) on fungus-growing ants. I have discussed this system previously, most recently in a draft chapter for a proposed book on Darwinian Agriculture. (Discussions at the meeting with a publisher were encouraging, but I need to prepare a more-detailed proposal.) The ants use antibiotics, produced by specific bacteria, to control parasitic fungi that threaten the fungi they grow for food. I have always wondered why mutant bacteria that don't produce these costly antibiotics don't take over. There's a long-term collective benefit to bacteria from living in a healthy ant colony, but natural selection is notoriously blind to the long term and to collective benefits.
So I was very interested in Cameron's report that Ainslie Little has found a yeast fungus that attacks the bacteria. I suggested that defense against yeast might provide an individual benefit to the bacteria from producing antibiotics, with defense against other fungi as a side effect, but Cameron says the yeasts have typically evolved resistance to the antibiotics produced by the bacteria in a given ant nest. It would be interesting to track antibiotic production and resistance over time, to see if there could be an arms race in progress. He says transfer of bacteria from ant to ant seems to be carefully controlled. Can the ants tell which bacteria are most effective at producing antibiotics?
After the meeting, my wife joined me for a weekend at Point Lobos, where we saw lots of lichens (another classic example of cooperation between species) and begging in cormorants and gulls, and the Monterey Bay Aquarium. We had dinner twice at Sea Harvest in Moss Landing, a good place for fresh local (and nonlocal) fish and great views of sea otters from their deck.