Supplementary Materials

3 downloads 0 Views 528KB Size Report
Francisco Rodríguez-Trelles, Rosa Tarrío, and Mauro Santos. Supplementary Methods and Data. Collection dates. Mount Pedroso and Berbikiz were sampled ...
Electronic Supplementary Material and Data rsbl.2013.0228

Genome-wide evolutionary response to a heat wave in Drosophila Francisco Rodríguez-Trelles, Rosa Tarrío, and Mauro Santos

Supplementary Methods and Data Collection dates Mount Pedroso and Berbikiz were sampled approximately in the same calendar date in the four seasons (spring, early summer, late summer and autumn) and the two years (2001 and 2012), except when adverse weather conditions prevented flies’ collections. Differences in collection dates between locations were kept to a minimum by sampling one right after the other during the same field trip. The dates of collections were as follows. Year 2011 spring: April 22 at Mount Pedroso and April 18 at Berbikiz. Year 2011 early summer: July 4 at Mount Pedroso and July 8 at Berbikiz. Year 2011 late summer: August 26 at Mount Pedroso and August 31 at Berbikiz. Year 2011 autumn: November 24 at Mount Pedroso and November 13 at Berbikiz. Year 2012 spring: May 11 at Mount Pedroso and May 14 at Berbikiz. Year 2012 early summer: July 4 at Mount Pedroso and July 7 at Berbikiz. Year 2012 late summer: August 26 at Mount Pedroso and August 28 at Berbikiz. Year 2012 autumn: November 12 at Mount Pedroso and November 14 at Berbikiz.

Genome-wide warm dose The chromosome arrangement frequencies are given in supplementary Table S1. Comprehensive summaries of the wealthy inversion polymorphism in D. subobscura revealed that the complex gene arrangements (with overlapping and nonoverlapping inversions) from more equatorial Palearctic populations are gradually replaced by the socalled standard gene arrangements in the five major acrocentric chromosomes as populations approach high latitudes [1, 2]. Therefore, we have estimated the genome-wide warm dose as

1

Electronic Supplementary Material and Data rsbl.2013.0228

the unweighted average of one minus the frequency of standard gene arrangements on each chromosome (the quantitative conclusions remain the same if we use the weighted average).

Genome-wide chromosome index An alternative way of analyzing the seasonal data at Mount Pedroso and Berbikiz is to perform a principal component analysis (PCA) on the centred (without scaling) inversion frequencies on all 16 (population × season × year) samples as in [3]. The figure for PC1, which accounts for 65.9% of the variance, against sample date is shown below (red lines: 2011; blue lines: 2012).

0.5

Mt. Pedroso

PC1

0.25

0 Berbikiz -0.25

-0.5

S

ES

LS

Season

A

The pattern for PC1 is clearly similar to that obtained for the genome-wide warm dose plotted in figure 2 of the paper (Friedman’s ANOVA on PC1 gives the same statistically

2

Electronic Supplementary Material and Data rsbl.2013.0228

significant seasonal pattern; 32  11.10 , P = 0.011). Interpretation of PC1 can be obtained from the PC coefficients (loadings): EST JST O3+4 OST AST O3+4+8 U1+2 A1 UST E1+2 O3+4+16+2 A1+2 U1 O7 O3+4+22 U2 O3+4+1 O3+4+2 E8 E1+2+9 U1+2+8 E1+2+9+3 A2 J1 E1+2+9+12 O3+4+7

-0.46573 -0.28707 -0.21533 -0.21510 -0.17736 -0.14962 -0.08052 -0.02614 -0.02418 -0.01294 -0.01068 -0.00079 0.00095 0.00110 0.00677 0.01146 0.01314 0.01564 0.01736 0.01750 0.09256 0.11611 0.20418 0.25572 0.32911 0.55455

These loadings reflect the pattern in figure 2: a weighted contrast between all standard arrangements against most ―warm-climate‖ inversions. An exception is warm-climate O3+4 , which attains a relatively high negative loading. O3+4 is known to be the weakest seasonal arrangement on chromosome O at Mount Pedroso and, at the same time, is the arrangement with the most pronounced long-term trend of increasing frequency [4]. This trend is also clear in our present samples at Mount Pedroso, and the PC loading for O3+4 cannot distinguish between a true seasonal pattern and/or a trend of increasing frequency. Comparatively, the warm dose in figure 2 gives the same pattern as PC1 but it is probably more intuitive and easy to grasp.

3

Electronic Supplementary Material and Data rsbl.2013.0228

Estimate of selection coefficients To roughly estimate the selection coefficients that could explain the increase of O3+4+7 after the heat wave we will assume here (i) a diallelic locus (i.e., O_ includes all gene arrangements but O3+4+7 on chromosome O) in a random mating population, (ii) that selection is acting on the egg-to-adult viability component, and (iii) that the frequency before selection (i.e., the frequency in zygotes during the heat wave) for gene arrangement O3+4+7 corresponds to the average spring frequency for the historical records at Mount Pedroso (p = 0.4009), which is basically similar to the average frequency in autumn (0.3487). Assuming also the relative fitness of karyotypes O3+4+7 / O3+4+7, O3+4+7 / O_, and O_ / O_ to be 1 , 1  hs , and 1  s , respectively, where s is a positive constant and 0  h  1 is the degree of dominance, the selective coefficients (relative survival rates) for different values of h can be estimated from the change in frequency of O3+4+7 p  p  p , where p  0.6429 is the frequency of O3+4+7 in spring 2011 after the heat wave. The estimates are as follows:

Selective value of Condition

O3+4+7 / O3+4+7 O3+4+7 / O_ O_ / O_ Change in p  p 

Estimate of s

—————————————— Dominance

1

1 s

1

 h  0 Additivity

1

 h  1 2 Recessiveness

1

1 1 s 2

1 s

1 s

1 s

 h  1

sq 2 p 1  sq 2

1 spq 2 1  sq

sqp 2 1  2sq  sq 2

1.05

0.91

0.81

Although there is a lot of uncertainty in these estimates, the selection coefficients are large enough as to critically question the reliability of this simple model (the dominance situation

4

Electronic Supplementary Material and Data rsbl.2013.0228

seems to be clearly unrealistic). There are at least two additional complications to be considered: gene flow and/or a more complex pattern of selection. Gene flow could have produced the genetic anomaly we observed in April 2011 assuming that any source population within migration distance from Mount Pedroso had sent a high number of O3+4+7 migrants. To test this possibility, we have revisited the available historical samples on inversion frequencies from sampling sites (i) within a radius equivalent to the distance between Mount Pedroso and Berbikiz (600 km), and (ii) at a higher distance than this radius from Mount Pedroso. In the first case some active dispersal could have happened, but in the second case we can rule out ―massive‖ active dispersal because Mount Pedroso and Berbikiz were genetically differentiated in all samples. The complication we found is that specific collection dates are not available for all samples; this is important because of the cyclical seasonal pattern of O3+4+7. In any case, the highest frequency of O3+4+7 in the 7 populations within the Mount Pedroso – Berbikiz radius was recorded at Pontedeume (79 km from Mount Pedroso) sampled in summer 1971: 0.741 [5]. The difference with Mount Pedroso is not very large when considering the average (0.557) and the upper limit (0.691) of historical frequencies of O3+4+7 at Mount Pedroso in the same season: Pontedeume deviates +1.9 from the historical summer average and +0.5 from the maximum frequency at that season; both values are within 2 (frequencies were arcsin-transformed). When considering the 10 populations outside the 600 km radius from Mount Pedroso, only Ronda (1,000 km) had a higher frequency of O3+4+7 than Mount Pedroso: 0.735 [6]. Although no information on specific collection dates was provided, this frequency is lower than that at Pontedeume. To sum up, we think that gene flow can be rule out as the (main) factor that could explain the genetic anomaly observed after the heat wave and that selection was the probable cause. A problem with the previous estimates of selection coefficients is that the model assumes that all selection occurs in the egg-to-adult viability (―early fitness‖) component, but

5

Electronic Supplementary Material and Data rsbl.2013.0228

it is quite likely that adult (―late‖) fitness components (e.g., female fertility and male mating success) were also affected by the heat wave. As already pointed out in the main text, the heat wave began on April 1, approximately three weeks before the collection dates in spring 2011. The average temperature in this 3-week period at Mount Pedroso was 16.4 C, with an average Tmax = 22.3 C. The thermal conditions during the heat wave were warm enough to allow one offspring generation because the egg-to-adult developmental time at these temperatures is about two-three weeks in D. subobscura [7]. If we assume that the intensity of selection was about the same for late (i.e., for the parentals that gave rise to the offspring generation we sampled) and early fitness components, p in the numerical exercise above should be halved and s  0.6 for all values of the degree of dominance h . This estimate might still look quite high, but it is worth noting that relative fitness estimates for mating success comparing OST / OST and O3+4+7 / O3+4+7 karyotypes can be as high as 0.33 under ―optimal‖ (17C) thermal conditions [8], and that ―cold-adapted O chromosomes‖ experience a loss in relative net fitness of 33% when placed in a warm environment at constant temperature of 22 C [9]. To sum up, it seems safe to conclude that selection coefficients during the heat wave were huge enough to (at least partially) explain the genetic anomaly in the frequency of O3+4+7 at Mount Pedroso.

6

Electronic Supplementary Material and Data rsbl.2013.0228

Table S1. Gene arrangement frequencies in spring (S), early summer (ES), late summer (LS) and autumn (A) at Mount Pedroso and Berbikiz in 2011 and 2012. Mount Pedroso 2011

2012

S

ES

LS

A

S

ES

LS

A

AST A1 A2 A1+2 N

0.247 0.014 0.740 0.000 73

0.182 0.000 0.818 0.000 55

0.241 0.000 0.759 0.000 87

0.269 0.015 0.716 0.000 134

0.261 0.018 0.721 0.000 111

0.255 0.043 0.702 0.000 47

0.204 0.010 0.786 0.000 103

0.263 0.000 0.737 0.000 118

EST E1+2 E1+2+9 E1+2+9+3 E1+2+9+12 E8 N

0.403 0.039 0.065 0.273 0.130 0.091 77

0.393 0.018 0.054 0.036 0.464 0.036 56

0.345 0.035 0.058 0.138 0.425 0.000 87

0.552 0.090 0.015 0.067 0.269 0.008 134

0.487 0.081 0.081 0.072 0.261 0.018 111

0.330 0.170 0.000 0.032 0.468 0.000 94

0.333 0.181 0.019 0.057 0.391 0.019 105

0.670 0.110 0.009 0.042 0.152 0.017 118

JST J1 N

0.234 0.766 77

0.228 0.772 57

0.184 0.816 87

0.373 0.627 134

0.405 0.595 111

0.309 0.692 94

0.257 0.743 105

0.390 0.610 118

OST O7 O3+4 O3+4+8 O3+4+16+2 O3+4+1 O3+4+2 O3+4+7 O3+4+22 N

0.014 0.007 0.186 0.064 0.007 0.029 0.014 0.643 0.036 140

0.019 0.019 0.150 0.084 0.000 0.000 0.056 0.673 0.000 107

0.023 0.012 0.250 0.023 0.000 0.023 0.023 0.622 0.023 172

0.049 0.011 0.372 0.143 0.011 0.000 0.026 0.350 0.038 266

0.069 0.018 0.262 0.151 0.009 0.009 0.018 0.399 0.064 218

0.075 0.011 0.362 0.106 0.011 0.000 0.011 0.383 0.043 94

0.039 0.010 0.280 0.097 0.005 0.019 0.034 0.483 0.034 207

0.051 0.013 0.438 0.140 0.013 0.013 0.026 0.255 0.051 235

UST U1 U2 U1+2 U1+2+8 N

0.000 0.000 0.000 0.766 0.234 77

0.000 0.000 0.000 0.821 0.179 56

0.000 0.000 0.024 0.831 0.145 83

0.000 0.000 0.008 0.836 0.157 134

0.000 0.000 0.018 0.856 0.126 111

0.011 0.011 0.000 0.798 0.181 94

0.010 0.000 0.010 0.771 0.210 105

0.017 0.000 0.000 0.822 0.161 118

7

Electronic Supplementary Material and Data rsbl.2013.0228

Table S1. (Continued.) Berbikiz 2011

2012

S

ES

LS

A

S

ES

LS

A

AST A1 A2 A1+2 N

0.253 0.010 0.727 0.010 99

0.172 0.022 0.806 0.000 93

0.223 0.029 0.748 0.000 103

0.338 0.042 0.620 0.000 142

0.426 0.040 0.535 0.000 101

0.273 0.030 0.697 0.000 66

0.253 0.000 0.747 0.000 87

0.336 0.009 0.655 0.000 113

EST E1+2 E1+2+9 E1+2+9+3 E1+2+9+12 E8 N

0.551 0.061 0.071 0.112 0.153 0.051 98

0.532 0.096 0.032 0.053 0.245 0.043 94

0.447 0.117 0.058 0.029 0.350 0.000 103

0.739 0.042 0.049 0.028 0.127 0.014 142

0.683 0.069 0.059 0.010 0.149 0.030 101

0.591 0.144 0.000 0.038 0.220 0.008 132

0.444 0.128 0.060 0.034 0.325 0.009 117

0.673 0.106 0.000 0.035 0.177 0.009 113

JST J1 N

0.337 0.663 98

0.305 0.695 95

0.291 0.709 103

0.444 0.634 153

0.495 0.505 101

0.303 0.697 132

0.333 0.667 117

0.389 0.611 113

OST O7 O3+4 O3+4+8 O3+4+16+2 O3+4+1 O3+4+2 O3+4+7 O3+4+22 N

0.156 0.005 0.312 0.129 0.005 0.022 0.038 0.328 0.005 186

0.083 0.017 0.365 0.099 0.011 0.017 0.044 0.359 0.006 181

0.109 0.015 0.416 0.089 0.010 0.040 0.010 0.297 0.015 202

0.241 0.004 0.394 0.181 0.007 0.007 0.025 0.131 0.011 282

0.181 0.015 0.352 0.191 0.015 0.000 0.015 0.221 0.010 199

0.273 0.030 0.250 0.136 0.000 0.000 0.015 0.273 0.023 132

0.112 0.012 0.394 0.118 0.006 0.000 0.006 0.324 0.029 170

0.142 0.013 0.385 0.168 0.013 0.009 0.013 0.243 0.013 226

UST U1 U2 U1+2 U1+2+8 N

0.000 0.000 0.000 0.816 0.184 98

0.010 0.000 0.021 0.854 0.115 96

0.010 0.010 0.000 0.874 0.107 103

0.014 0.000 0.000 0.880 0.106 142

0.030 0.000 0.000 0.901 0.069 101

0.030 0.008 0.000 0.705 0.258 132

0.034 0.000 0.000 0.821 0.145 117

0.018 0.000 0.000 0.858 0.124 113

8

Electronic Supplementary Material and Data rsbl.2013.0228

References for Supplementary Information 1 2

3 4

5 6

7

8 9

Krimbas CB, Loukas M. 1980 The inversion polymorphism of Drosophila subobscura. Evol. Biol. 12, 163-234. Menozzi P, Krimbas CB. 1992 The inversion polymorphism of Drosophila subobscura revisited: synthetic maps of gene arrangement frequencies and their interpretation. J. Evol. Biol. 5, 625-641. Balanyà J, Oller JM, Huey RB, Gilchrist GW, Serra L. 2006 Global genetic change tracks global climate warming in Drosophila subobscura. Science 313, 1773-1775. Rodríguez-Trelles, F, Alvarez G, Zapata C. 1996 Time-series analysis of seasonal changes of the O inversion polymorphism of Drosophila subobscura. Genetics 142, 179-187. de Frutos R. 1972. Contribution to the study of chromosomal polymorphism in the Spanish populations of Drosophila subobscura. Genét. Ibér. 24, 123-140. Garcia MP, Prevosti A. 1981 Association between allozyme alleles and chromosomal arrangements of the O chromosome in Drosophila subobscura. I. Data of natural populations. Genét. Ibér. 33, 151-174. Santos M, Fernández Iriarte P, Céspedes W, Balanyà J, Fontdevila A, Serra L. 2004 Swift laboratory thermal evolution of wing shape (but not size) in Drosophila subobscura and its relationship with chromosomal inversion polymorphism. J. Evol. Biol. 17, 841-855. Santos M, Tarrío R, Zapata C, Alvarez G. 1986 Sexual selection on chromosomal polymorphism in Drosophila subobscura. Heredity 57, 161-169. Santos M. 2007 Evolution of total net fitness in thermal lines: Drosophila subobscura likes it "warm". J. Evol. Biol. 20, 2361-2370.

9