Figure 1.
TMV Decomposes Reaction Norm Shape Variation into Three Modes of Variation
Vertical shifts produce changes in average performance, horizontal shifts produce changes in the optimal temperature, and generalist-specialist variation produces changes in niche width. An identical template polynomial is shown in each chart as a black dashed reaction norm. The transitions from blue to red curves illustrate hypothetical contributions of the three modes of variation to adaptation to high temperature.
Figure 2.
Holder and Bull [22] used a clone of the bacteriophage G4 to found a population that was propagated through 50 serial transfers at 41.5 °C, followed by 50 serial transfers at 44 °C. In each serial transfer 1 × 106 phages were added to a 10 ml culture containing 1 × 109 exponentially growing E. coli hosts, incubated for approximately 45 min at the specified temperature, and then treated with chloroform to kill any remaining hosts. During the 45-min incubation, the phage population size increased to approximately 1 × 1010 phages, and 1 × 106 of these phages were used to initiate the next serial transfer. The only exception occurred after the 50th serial transfer, when the population was forced through a single phage bottleneck to homogenize the population. In the current study, we isolated single phage from this evolving population at transfers 20, 50, and 100 (designated G420, G450, and G4100). We characterized the performance of these phages and the ancestral G4 (designated G40) across a range of temperatures.
Figure 3.
Correlated Responses to Adaptation at High Temperature
(A) Reaction norms estimated using TMV explain 71.79% of the sums of squares in the data. Data are mean growth rates ± s.e.m. for each phage genotype determined from three or four replicate measures at each temperature. (B) Correlated responses to selection at 41.5 °C (transfers 1–50; open circles) and to selection at 44 °C (transfers 51–100; closed circles). Data are the differences in mean ln(growth rate) ± s.e.d. between G450 and G40, and between G4100 and G450, respectively.
Figure 4.
Modes of Variation among the Reaction Norms of Evolving Genotypes
TMV was used to decompose the variation among the reaction norms described in Figure 3 into three modes of variation. Each uppermost graph depicts the values of the fitted parameters from the model and the bottom graphs illustrate the impact of each parameter, if considered alone, on the shape of the reaction norm of G40. lnW is the natural log of growth rate.
Table 1.
Decomposition of Shape Variation
Table 2.
Genetic Basis of Adaptation to High Temperature
Figure 5.
Growth Rates of Phage Isolated from Nature at High Temperatures
(A) Phylogenetic relationships of G4-like phages recently isolated from environmental samples, modified from [24]. The genotypes of two pairs of sister taxa that differ at a locus (gene H: A47V) determined to have a large effect on reaction norm shape are indicated with blue (alanine) and red (valine). (B and C) Thermal reaction norms of the two phage pairs. Data are means ± s.e.m. determined from four or five replicate measures at each temperature.
Table 3.
ANOVA Describing the Effect of Gene H Mutation A47V in Natural Phage Isolates