A series of competition experiments with two genotypes of Escherichia coli showed that each genotype was favored when it was the minority, allowing their coexistence in a stable polymorphism. In these experiments, glucose was the sole source of carbon provided, and its concentration was limiting to population density. Thus, the stable polymorphism did not conform to a simple model of competitive exclusion. In similar experiments also with glucose as the sole resource, we considered two hypotheses that might explain the observed coexistence: (1) a strictly demographic trade—off, such that one genotype is competitively superior when glucose is abundant whereas the other genotype is the better competitor for sparse glucose; and (2) a cross—feeding interaction, whereby the superior competitor for glucose excretes a metabolite that acts as a second resource for which the other genotype is the better competitor. Although there was a demographic tradeoff, the advantage to the superior competitor at high glucose concentrations was too large (given the initial concentration of glucose used in these experiments) to allow the second genotype to invade when rare at the observed rate. Therefore, the second genotype must have had some other advantage that allowed it to readily invade a population of the superior competitor for glucose. Indeed, the second genotype could increase in abundance after glucose was depleted, but only in the presence of the superior competitor for glucose, thus implicating a cross—feeding interaction. These results confirmed earlier studies showing that populations of E. coli can maintain ecologically relevant genetic diversity even in a simple environment.

Many bacterial species exhibit strong linkage disequilibrium of their chromosomal genes, which apparently indicates restricted recombination between alleles at different loci. The extent to which restricted recombination reflects limited migration between geographically isolated populations versus infrequent mixis of genotypes within populations is more difficult to determine. We examined the genetic structure of Rhizobium kguminosarum biovar phaseoli populations associated with wild and cultivated beans (Phaseolus spp.) over several spatial scales, ranging from individual host plants to throughout the Western Hemisphere. We observed significant linkage disequilibrium at scales at least as small as a cultivated plot. However, the amount of disequilibrium was much greater among isolates collected throughout the Western Hemisphere than among isolates from one area of Mexico, even when disequilibrium was quantified using an index that scales for allelic diversity. This finding suggests that limited migration between populations contributes substantially to linkage disequilibrium in Rhizobium. We also compared the genetic structure for R. leguminosarum bv. phaseoli taken from a cultivated plot with that for Escherichia colt obtained from one human host in an earlier study. Even at this frne scale, linkage disequilibrium in E. coil was very near the theoretical maximum level, whereas it was much less extreme in the local population of Rhizobium. Thus, the genetic structure for R. leguminosarum bv. phaseoli does not exclude the possibility of frequent mixis within local populations.

No abstract

Current knowledge of genotypic and phenotypic diversity in the species Escherichia coli is based almost entirely on strains recovered from humans or zoo animals. In this study, we analyzed a collection of 202 strains obtained from 81 mammalian species representing 39 families and 14 orders in Australia and the Americas, as well as several reference strains; we also included a strain from a reptile and 10 from different families of birds collected in Mexico. The strains were characterized genotypically by multilocus enzyme electrophoresis (MLEE) and phenotypically by patterns of sugar utilization, antibiotic resistance, and plasmid profile. MLEE analysis yielded an estimated genetic diversity (H) of 0.682 for 11 loci. The observed genetic diversity in this sample is the greatest yet reported for E. coli. However, this genetic diversity is not randomly distributed; geographic effects and host taxonomic group accounted for most of the genetic differentiation. The genetic relationship among the strains showed that they are more associated by origin and host order than is expected by chance. In a dendrogram, the ancestral cluster includes primarily strains from Australia and ECOR strains from groups B and C. The most differentiated E. coli in our analysis are strains from Mexican carnivores and strains from humans, including those in the ECOR group A. The kinds and numbers of sugars utilized by the strains varied by host taxonomic group and country of origin. Strains isolated from bats were found to exploit the greatest range of sugars, while those from primates utilized the fewest. Toxins are more frequent in strains from rodents from both continents than in any other taxonomic group. Strains from Mexican wild mammals were, on average, as resistant to antibiotics as strains from humans in cities. On average, the Australian strains presented a lower antibiotic resistance than the Mexican strains. However, strains recovered from hosts in cities carried significantly more plasmids than did strains isolated from wild mammals. Previous studies have shown that natural populations of E. coliharbor an extensive genetic diversity that is organized in a limited number of clones. However, knowledge of this worldwide bacterium has been limited. Here, we suggest that the strains from a wide range of wild hosts from different regions of the world are organized in an ecotypic structure where adaptation to the host plays an important role in the population structure.