Ten years of life in compost: temporal and spatial variation of North German Caenorhabditis elegans populations

Abstract

The nematode Caenorhabditis elegans is a central laboratory model system in almost all biological disciplines, yet its natural life history and population biology are largely unexplored. Such information is essential for in-depth understanding of the nematode's biology because its natural ecology provides the context, in which its traits and the underlying molecular mechanisms evolved. We characterized natural phenotypic and genetic variation among North German C. elegans isolates. We used the unique opportunity to compare samples collected 10 years apart from the same compost heap and additionally included recent samples for this and a second site, collected across a 1.5-year period. Our analysis revealed significant population genetic differentiation between locations, across the 10-year time period, but for only one location a trend across the shorter time frame. Significant variation was similarly found for phenotypic traits of likely importance in nature, such as choice behavior and population growth in the presence of pathogens or naturally associated bacteria. Phenotypic variation was significantly influenced by C. elegans genotype, time of isolation, and sampling site. The here studied C. elegans isolates may provide a valuable, genetically variable resource for future dissection of naturally relevant gene functions.

Keywords: Bacillus thuringiensis; Caenorhabditis elegans; Serratia; microsatellites; natural variation; population genetics.

Figure 1

Microsatellite analysis of the North German populations across time and space. (A) Minimum spanning network of the genotype relationships. The strains N2 and CB4856 were included as a reference. The network was inferred using F_ST as a measure for genetic distance. Branch lengths correspond to the number of different alleles (see scale in legend) and circle sizes correspond to the number of isolates per genotype. Dashed lines indicate alternative connections between genotypes. The Kiel genotypes are given in different shades of purple depending on their occurrence across the time periods where light purple includes genotypes either found in only late 2011 and/or one or both of the later time periods; the genotypes that were unique in mid-2012 and late 2012 are given in purple and dark purple, respectively. Green color indicates genotypes from Roxel. (B) Occurrence of genotypes within Kiel across time. Genotype g4 in late 2012 appeared as one allele combination in the heterozygous genotype g15 and is thus given in light gray. (C) Genetic differentiation of subpopulations as F_ST values where 0 indicates the absence of differentiation (see right scale). The long-term comparison between Roxel 2002 and all subpopulations of Roxel 2011/12 and Kiel 2011/12 showed significant differences (all P < 0.001). All 2011/12 subpopulations from Kiel versus Roxel differed significantly from each other (all P < 0.001). No significant difference was found in the short-term comparisons among the subpopulations from Roxel 2011/12, whereas there was a trend for differences among the subpopulations from Kiel 2011/12 (all P < 0.1).

Figure 2

Population growth of natural Caenorhabditis elegans isolates. The population growth rate of the strains from Roxel (2002 and 2012) and Kiel (2012) and the wild-type strain N2 on (A) Escherichia coli, (B) Serratia sp., and (C) Serpens sp. was analyzed after 5 days and is shown as mean population growth per initial worm relative to the mean population growth of N2 on OP50 per initial worm (indicated as dashed line). The error bars indicate standard error of the mean. Genotype numbers below strain designations refer to those from Table S2.

Figure 3

Population growth of natural Caenorhabditis elegans isolates. The population growth rate of the strains from Roxel (2002 and 2012) and Kiel (2012) and the wild-type strain N2 was analyzed on (A) Bacillus thuringiensis DSM350 and (B) BT247 after 5 days and is shown as mean population growth per initial worm relative to the mean population growth of N2 on DSM350 per initial worm (indicated as dashed line). The error bars indicate standard error of the mean. Note that, the scales of DSM350 and BT247 differ. Genotype numbers are given below strain codes and are identical to those in Table S2.

Figure 4

Choice behavior of natural Caenorhabditis elegans isolates. The choice behavior of strains from Roxel (2002 and 2012) and Kiel (2012) and the wild-type strain N2 was analyzed on Serratia sp. (A), Serpens sp., (B) and BT247 (C) after 14 and 24 h. The bars show medians and the error bars median absolute deviation (mad). Genotype numbers are given below strain designations and are identical to those from Table S2.

Figure 5

Correlation between population growth rate and choice index on Serratia sp. The natural Caenorhabditis elegans isolates were collected in 2012 in Kiel (purple dots) and Roxel (green dots). Lines are predicted from a linear model. Shaded areas indicate the 95% confidence interval. Error bars denote standard error of the mean.