Portal de Periódicos da CAPES

Viabilities of Drosophila subobscura homo- and heterokaryotypes at optimal and stress temperatures. I. Analysis over several years

2010; BioMed Central; Volume: 147; Issue: 2 Linguagem: Inglês

10.1111/j.1601-5223.2009.2163.x

1601-5223

Goran Zivanovic, Francesc Mestres,

Genetic and Environmental Crop Studies

HereditasVolume 147, Issue 2 p. 70-81 Open Access Viabilities of Drosophila subobscura homo- and heterokaryotypes at optimal and stress temperatures. I. Analysis over several years Goran Zivanovic, Goran Zivanovic Department of Genetics, Institute for Biological Research "Sinisa Stankovic" University of Belgrade, SerbiaSearch for more papers by this authorFrancesc Mestres, Francesc Mestres Department of Genetics, Institute for Biological Research "Sinisa Stankovic" University of Belgrade, SerbiaSearch for more papers by this author Goran Zivanovic, Goran Zivanovic Department of Genetics, Institute for Biological Research "Sinisa Stankovic" University of Belgrade, SerbiaSearch for more papers by this authorFrancesc Mestres, Francesc Mestres Department of Genetics, Institute for Biological Research "Sinisa Stankovic" University of Belgrade, SerbiaSearch for more papers by this author First published: 04 May 2010 https://doi.org/10.1111/j.1601-5223.2009.2163.xCitations: 2 Goran Zivanovic, Department of Genetics, Institute for Biological Research "Sinisa Stankovic", University of Belgrade, Bulevar Despota Stefana 142, 11000 Belgrade, Serbia. E-mail: goranziv@ibiss.bg.ac.rs AboutSectionsPDF ToolsExport citationAdd to favoritesTrack citation ShareShare Give accessShare full text accessShare full-text accessPlease review our Terms and Conditions of Use and check box below to share full-text version of article.I have read and accept the Wiley Online Library Terms and Conditions of UseShareable LinkUse the link below to share a full-text version of this article with your friends and colleagues. Learn more.Copy URL Share a linkShare onFacebookTwitterLinked InRedditWechat Abstract The interactions of lethal and non-lethal genes and their contributions to the viability of Drosophila inversion karyotypes are not well understood. This is especially true under variable environmental conditions. Here we examine the viability of natural chromosomal O-inversion homo- and heterokaryotypes in a D. subobscura population from Avala Mountain, Serbia. The observations we report were performed at a range of temperatures over several years. The heterotic effect of O-lethal heterozygotes on viability was found to be independent of the effects of inversion backgrounds and temperature. Positive epistatic interactions of lethal, mildly deleterious (subvital) and quasinormal (normal) genes were found in O-inversions in heterokaryotypes but not in homokaryotypes. These interactions were independent of temperature. This finding could explain the limitation of the genetic load in D. subobscura populations. In the population analyzed, annual fluctuations in the frequencies of certain chromosomal arrangements, karyotypes and non-lethal chromosomes under cold-stress temperatures seemed to indicate a correlation between these polymorphisms and environmental conditions. Our results indicate that there is a response in tolerance to extreme temperatures that may be due to natural selection. The differences in mean viability between some O-inversion karyotypic combinations indicate that there are differences in their tolerance to variable temperatures. All our results suggest that both frequency-dependent and supergene selection are mechanisms that protect O-chromosomal inversions. Chromosomal inversions may be genetically differentiated and coadapted complexes in D. subobscura populations. Many animals and plants are able to adapt rapidly to changes in environmental conditions (Endler 1986). Chromosomal inversion polymorphism in Drosophila, an extensively studied system, is a sensitive indicator of the effects of changes in environmental conditions on genetic composition (reviewed by Sperlich and Pfriem 1986; Krimbas and Powell 1992; Powell 1997). Chromosomal arrangements in Drosophila are considered to be coadapted complexes whose most prominent effect is the suppression of gene recombination in heterokaryotypic combinations (Dobzhansky 1970). This indicates that inversions are involved in maintaining the linkage between epistatically interacting coadapted alleles (Soule 1973). Consequently, lethal and non-lethal genes are ideal markers for studying this kind of interaction and the results may be of great importance for evolutionary biology (Kondrashov 1993) and conservation genetics (Hedrick 2001). Although the interaction of lethal and non-lethal genes in natural inversion karyotypes could have relevant consequences at a population level, few studies have addressed this topic in Drosophila. Consequently there is a lack of comparative results regarding viability between genetically heterozygous natural inversion structural homo- and heterokaryotypes (Zivanovic and Marinkovic 2003). Indeed, very few studies have addressed the effects of such interactions on viability in natural inversion karyotypes under optimal conditions and at stress temperatures in natural populations of Drosophila. Drosophila subobscura is characterized by abundant chromosome inversion polymorphism. The O-chromo-some is the longest and most polymorphic (Krimbas 1992, 1993), and can be analyzed in depth using the lethal balancer strain Va/Ba, the only one available in this species (Sperlich et al. 1977). As the O-chromosome is homologous to the 3R chromosome arm of D. melanogaster and the second chromosome of D. pseudoobscura, it is ideal for comparative purposes. Finally, as genes for heat-stress proteins are concentrated in the O3+4 region of the O-chromosome in D. subobscura (Moltó et al. 1992), we chose this chromosome to test possible associations of O-chromosomal inversions as well as O-inversion karyotypes with distinct tolerance to cold and heat. Testing these associations, especially in samples observed over several years with evident climatic changes, provides valuable information on the adaptive capacity of O-chromosomal inversion and genetic load polymorphisms in natural populations. The effects of climate changes on D. subobscura chromosomal polymorphism has been studied previously (Solé et al. 2002; Balanyá et al. 2004, 2006), but as far as we know, never in conjunction with viability. Here we compare the viability of several genetically heterozygous natural chromosomal O-inversion homo-and heterokaryotypes in a D. subobscura population collected from the Avala Mountain near Belgrade, Serbia. Populations from southeastern Europe are especially interesting because the region is considered to have been a continental refuge for both animals and plants during the last glaciation (Taberlet et al. 1998; Hewitt 1999, 2000; Heckel et al. 2005). Viability was studied over several years and, as in recent research (Zivanovtc and Martnkovtc 2003), across a range of temperatures: optimal (nearly 19°C), cold (12°C, 14–15°C) and hot (25°C). Our study is focused on the heterotic effect of O-lethal heterozygotes on viability and aims to ascertain whether this effect is independent of the general genetic background and of temperature. We also analyze whether the expected positive epistatic interactions of lethal and non-lethal genes depend on distinct environmental temperatures over several years. Furthermore, we examine whether there was a non-random distribution of combinations of viability classes (with distinct combinations of genes) among O-inversion karyotypes. Our results may also provide additional support for the notion that chromosomal inversions in D. subobscura are adaptive. Finally, recent comparative results concerning the viability of natural O-inversion homo- and heterokaryotypes with distinct frequencies in D. subobscura populations from southeastern Europe point to several balancing selection mechanisms that protect chromosomal inversions as coadapted complexes (Zivanovic and Martnkovic 2003). These mechanisms are considered key elements in maintaining the chromosomal polymorphism of populations (Hoffmann and Rieseberg 2008) and therefore we examined them over several years and over a range of temperatures. MATERIAL AND METHODS Population and samples We analyzed a natural population of Drosophila subobscura sampled from Avala Mountain (44°48′N), approximately 18 km south of Belgrade, Serbia. Flies were collected from a forest with polydominant communities of Fagetum submontanum mixtum that is approximately 450 m a.s.l. The flies were sampled at the same site in September 2003, 2004 and 2005 (for simplicity these samples are hereafter referred to by year only). Analyses of chromosomal arrangements and viabilities The experimental design (based on Sperlich et al. 1977) is described in detail in Zivanovic and Martnkovic (2003) and presented in Fig. 1. From the sample population, only wild males and male offspring of wild females were used to analyze O chromosomal inversions and genetic load. In the P generation, males (usually 4–5 days old) were individually crossed with virgin females of the Ku (Kusnacht) strain. After several days, the same males were also individually crossed with virgin females of the balanced lethal strain Va/Ba to determine the genetic load. When the males analyzed were O-inversion homokaryotypic (O′/O′), ten or more Va males of the F1 generation (in many cases as many as twenty) were crossed individually with Va/Ba virgin females. In contrast, when the males used for crossbreeding were O-inversion heterokaryotypic (O′/O″), at least eight Va males of the F1 generation were individually crossed with virgin females of the Ku strain. After several days the same males (F1 generation) were individually crossed with virgin females of the Va/Ba strain. Following that, O-chromosomal inversions were analyzed to find the two O-arrangements in the O-inversion heterokaryotypic combination of the males. In the F2 generation, when the males analyzed were O-inversion homokaryotypic in each of (at least) ten crosses, three Va males and three virgin Va females were inter-crossed. It is worth to point out that the initial homokaryotypic males (O′/O′) were not homozygous and the homologous chromosomes are indicated by O′1 and O′2 (Fig. 1). This procedure was repeated twice more (to test viability under a range of temperatures), for each of (at least) ten crosses performed. Flies were kept in tubes with at least 20 ml of standard medium. After several days, they were placed separately in different tubes with the standard medium for each of the three cases analyzed. Next, each of these three tubes (with their contents of eggs and young larvae) was exposed to a controlled temperature of 12°C, 19°C and 25°C. This process (exposing each tube to a distinct temperature) was repeated twice, after every new putting of flies was done. This exposure experiment was performed on samples collected in 2003 and 2004. However, flies from the 2005 sample were kept at a non-controlled temperature which fluctuated between cold and almost optimal (with a mean temperature of about 14–15°C). To prevent the effects of overpopulation on larval growth, extra yeast was added to the cultures every few days. In the F2 generation, when the males analyzed were O-inversion heterokaryotypic for the two O-arrangements, three Va males and three Va females were crossed. This procedure was repeated twice to test viability at a cold (12°C), almost optimal (19°C), hot (25°C) and non-controlled (only for the 2005 sample) temperature. Furthermore, crosses between these two O-arrangements (Va males with the first arrangement crossed with Va females with the second) were carried out to test the viability of the reconstructed O-inversion heter-okaryotype, also under the same four temperature regimes. After several days, the flies from all the crosses were transferred to fresh tubes. This procedure was repeated and after every new putting flies, old tubes with eggs and young larvae contents were placed at 12°C, 19°C or 25°C or maintained in non-controlled conditions. In the F3 generation, when the males analyzed were O-inversion homokaryotypic for every of (at least) ten individual crosses, the proportion of wild-type individuals from crosses of Va males and Va females was determined at 12°C, 19°C and 25°C, except for the 2005 sample which was subjected to a non-controlled temperature. The average production of flies under optimal temperature, for all the crosses performed, was approximately 200. The larvae under cold stress developed markedly more slowly (at least two months) and fewer flies emerged (a mean of 80). The proportion of flies that hatched allowed us to establish a specific viability class. Later, when we had indirectly identified the arrangements in the O-inversion homokaryotypic flies, Va males of one O-arrangement were crossed with virgin Va females of another. As a result the O-inversion homokaryotype of the males analyzed was reconstructed in the next generation. In the F3 generation, when the males analyzed were O-inversion heterokaryotypic, we calculated the proportion of wild-type individuals for each O-arrangement and also reconstructed the O-inversion heterokaryotype for the different temperatures. In the F4 generation the proportion of wild-type individuals for reconstructed O-inversion homokaryotypes was determined under the four temperature conditions established. Figure 1Open in figure viewerPowerPoint Experimental design for viability and chromosomal analyses of homo- and heterokaryotypic combinations for the O chromosome of D. subobscura at different temperatures. Fly samples were collected at different years from Avala Mountain population. In the F2 generation of the experimental design twenty crosses (15 for 2003 and 5 for 2004) were carried out at optimal temperature, between Va males and Va females with distinct wild O-chromosomes and independently of their inversion backgrounds. We then calculated the proportion of wild-type individuals in the F3 generation, and the criterion of Dobzhansky and Spassky (1963) and Lewontin (1974) was used to establish viability classes. The mean viability of 20 crosses of distinct wild O-chromosomes from our sample population was used to establish viability classes for the 2004 and 2005 samples. Lethal genes on the O-chromosome were maintained in a heterozygous condition using the Va balancer chromosome. As the lethal balancer chromosome Va presents the naturally occurring O3+4 arrangement, recombination is not prevented in this region if the wild chromosome carries this arrangement with any other overlapped inversion (Loukas et al. 1980, Mestres et al. 1990). However, we recently presented some evidence that a possible recombination between Va chromosome and wild O3+4 chromosomes does not invalidate our results (Zivanovic and Marinkovic 2003). All strains, stocks and crosses were kept at 19°C, a near-optimal temperature for D. subobscura (optimal is about 18°C) (Krimbas 1993). Flies were fed a standard cornmeal-sugar-agar-yeast medium and kept at 60% relative humidity under a 12 h/12 h light/dark cycle. Statistical methods To measure the amount of chromosomal inversion polymorphism in the sample population, we used the index of free recombination (IFR). Comparisons between the mean viability of the O-inversion karyotypes and O-arrangements at the temperatures tested were computed by analysis of variance, using the arcsine transformation of the values (Sokal and Rohlf 1981). RESULTS Chromosomal polymorphism The Drosophila subobscura population from Avala Mountain showed eight chromosomal arrangements of the O-chromosome, and fourteen of the other four chromosomes (Table 1). The only difference in frequency was observed between 2003 and 2004 samples, where the Ost arrangement showed a higher frequency in 2003 (χ2= 2.95, DF=1, p < 0.10). A total of 34 distinct karyotypic combinations were found for the autosomes analyzed (Table 2). There were significant deviations from Hardy-Weinberg equilibrium for all three samples: 2003 (χ2= 108.9, DF = 21, p < 0.001), 2004 (χ2= 179.7, DF = 20, p < 0.001) and 2005 (χ2= 108.3, DF = 29, p < 0.001). These deviations are explained by an excess of some J-inversion (Jst/Jst). U-inversion (Ust/U1, Ust/U1+2, U1+2/U1+8+2, U1+2+6/U1+2+6), E-inversion (EstE1+2) and O-inversion (Ost/Ost, Ost/O3+4+6, O6/O3+4+6, O3+4+1/O3+4+5) karyotypic combinations. The IFR values were similar for all three samples: 2003 (83.32±2.45), 2004 (80.18±2.15) and 2005 (82.14±1.88). The pooled IFR of the three samples was 82.14±1.26. Table 1. Frequencies of chromosomal arrangements observed during different years in D. subobscura Avala Mountain population. September 2003 2004 2005 Chromosomal arrangements n % n % n % Ast 7 46.7 7 77.8 11 50.0 A1 3 20.0 2 22.2 7 31.8 A2 5 33.3 / / 4 18.2 Total 15 9 22 Jst 4 13.3 3 16.7 11 25.0 J1 26 86.7 15 83.3 33 75.0 Total 30 18 44 Ust 4 13.3 2 11.1 6 13.6 U1 / / / / 1 2.3 U1+2 15 50.0 7 38.9 20 45.5 U1+2+6 7 23.4 7 38.9 15 34.1 U1+8+2 4 13.3 2 11.1 2 4.5 Total 30 18 44 Est 8 26.7 7 38.9 18 40.9 E1+2 / / 1 5.5 1 2.3 E1+2+9 14 46.7 6 33.4 16 36.4 E8 8 26.6 4 22.2 9 20.4 Total 30 18 44 Ost 10 33.4 4 14.3 16 28.6 O6 / / 1 3.6 2 3.6 O3+4 14 46.7 16 57.1 23 41.0 O3+4+1 3 10.0 3 10.7 6 10.7 O3+4+5 1 3.3 / / 1 1.8 O3+4+6 1 3.3 2 7.1 1 1.8 O3+4+8 / / 1 3.6 1 1.8 O3+4+22 1 3.3 1 3.6 6 10.7 Total 30 28 56 Table 2. The frequencies of chromosomal karyotypes observed during different years in D. subobscura Avala Mountain population. September 2003 2004 2005 Chromosomal karyotypes n % n % n % Jst/Jst 1 6.7 / / 1 4.5 Jst/J1 2 13.3 3 33.3 9 40.9 J1/J1 12 80.0 6 66.7 12 54.6 Total 15 9 22 Ust/Ust / / / / 1 4.5 Ust/U1 / / / / 1 4.5 Ust/U1+2 2 13.3 2 22.2 2 9.1 Ust/U1+2+6 1 6.7 / / 1 4.5 Ust/U1+8+2 1 6.7 / / / / U1+2/U1+2 5 33.3 / / 3 13.7 U1+2/U1+2+6 / / 3 33.3 10 45.5 U1+2/U1+8+2 3 20.0 2 22.2 2 9.1 U1+2+6/U1+2+6 3 20.0 2 22.2 2 9.1 Total 15 9 22 Est/Est / / 1 11.1 3 13.7 Est/E1+2 / / 1 11.1 1 4.5 Est/E1+2+9 5 33.3 2 22.2 6 27.3 Est/E8 3 20.0 2 22.2 5 22.7 E1+2+9/E1+2+9 3 20.0 1 11.1 4 18.2 E1+2+9/E8 3 20.0 2 22.2 2 9.1 E8/E8 1 6.7 / / 1 4.5 Total 15 9 22 Ost/Ost 3 20.0 1 7.1 3 10.7 Ost/O3+4 2 13.3 1 7.1 5 17.8 Ost/O3+4+1 1 6.7 1 7.1 1 3.6 Ost/O3+4+6 1 6.7 / / 1 3.6 Ost/O3+4+22 / / / / 3 10.7 O6/O3+4 / / / / 2 7.1 O6/O3+4+6 / / 1 7.1 / / O3+4/O3+4 4 26.6 5 35.7 5 17.8 O3+4/O3+4+1 2 13.3 2 14.2 3 10.7 O3+4/O3+4+5 1 6.7 / / / / O3+4/O3+4+6 / / 1 7.1 / / O3+4/O3+4+8 / / 1 7.1 1 3.6 O3+4/O3+4+22 1 6.7 1 7.1 2 7.1 O3+4+1/O3+4+5 / / / / 1 3.6 O3+4+1/O3+4+22 / / / / 1 3.6 Total 15 14 28 Viabilities of chromosomal arrangements and temperature In a sample of 31 O-chromosome homozygotes at nearly optimal temperature (19°C), the viability distribution followed the usual bimodal pattern (Dobzhansky et al. 1977). Furthermore, a bimodal viability distribution was also observed in 16 and 21 homozygotes of O-chromosomes at 12°C and 25°C, respectively (2003 sample). In the two latter cases, the peak for quasinormal genes was lower due to an increase in the semilethal viability class under cold stress (12°C) and also a less common than expected normal viability class under both cold (12°C) and heat (25°C) stress. A similar result was obtained from 42 O-chromosome homozygotes under mild cold stress (14–15°C) (September 2005). In heterozygotes for several wild O-chromosomes, lethal and semilethal viability classes were absent at near optimal temperature only (2003 and 2004 samples). Furthermore, heterozygotes for wild O-chromosomes had a higher mean viability than homozygotes for O-chromosomes from the same samples, even when lethals were excluded from the calculations. The total genetic load, calculated following Crow (1958), was 0.4676 and 0.5193 for the 2003 and 2005 samples, respectively. In 2003, the mean number of lethals per chromosome (computed following Loukas et al. 1980) at 12°C, 19°C and 25°C was 0.371, 0.235 and 0.328, respectively. Finally, in the 2005 sample the mean number of lethals per chromosome was 0.371 (under non-controlled conditions and mean temperature 14–15°C). Viability of genetic combinations at different temperatures The viability classes of the chromosomes (L = lethal, SL = semilethal, SBV=subvital, N = normal) as well as the frequencies of the viability classes of genetic combinations (for instance: L/L = lethal/lethal, L/SL = lethal/semilethal) for natural O-inversion homo- and heterokaryotypes at different of temperatures are presented in Table 3a (by karyotype) and Table 3b (by year). The distributions of all viability classes for O-homozygous chromosomes derived from O-inversion heterokaryotypes at different temperatures were similar (χ2= 12.28, DF = 9, p > 0.05). When some classes were clustered, the viability of semilethals was grater than expected at 12°C (cold stress), but less than expected under mild cold stress (14–15°C) and nearly optimal (19°C) temperature (χ2= 4.30, DF=1, p < 0.05). The opposite result was observed for the normal viability class, which was more frequent than expected at 19°C and less frequent under cold-, mild cold- and heat-stress (χ2= 5.66, DF=l,p< 0.05). Table 3a. Frequencies of viability classes and combinations of viability classes in different O-inversion homo- and heterokaryotypes observed at different temperatures during the three different years. Homokaryotypes Heterokaryotypes 12°C 14–15°C 19°C 25°C 12°C 14–15°C 19°C 25°C Viability classes n % n % n % n % n % n % n % n % L 2 33.3 2 20.0 3 25.0 2 28.6 3 30.0 11 34.4 4 21.0 4 28.6 SL 3 50.0 2 20.0 3 25.0 3 42.8 4 40.0 4 12.5 2 10.5 3 21.4 SBV 1 16.7 2 20.0 1 8.3 1 14.3 / / 7 21.9 1 5.3 2 14.3 N / / 4 40.0 5 41.7 1 14.3 3 30.0 10 31.2 12 63.2 5 35.7 Total 6 10 12 7 10 32 19 14 Combinations of viability classes L/L / / / / / 1 33.3 / / 2 15.4 / / / / L/SL 1 50.0 / / 20.0 / / 1 20.0 2 15.4 1 12.5 2 28.6 L/SBV / / 1 33.3 / / / / / 1 7.7 / / / / L/N / / / / 20.0 / / 2 40.0 2 15.4 3 37.5 2 28.6 SL/SL / / / / / / / 1 20.0 / / / / / / SL/SBV 1 50.0 / / / 1 33.3 / / / / / / / / SL/N / / 2 66.6 20.0 1 33.3 1 20.0 1 7.7 1 12.5 1 14.3 SBV/SBV / / / / / / / / / 2 15.4 / / 1 14.3 SBV/N / / / / 20.0 / / / / 1 7.7 / / / / N/N / / / / 20.0 / / / / 2 15.4 3 37.5 1 14.3 Total 2 3 5 3 5 13 8 7 Table 3b. Frequencies of viability classes and combinations of viability classes in different O-inversion karyotypes observed at different temperature conditions during September of the three different years. September 2003 September 2004 September 2005 12°C 19°C 25°C 19°C 14–15°C Viability classes n % n % n % n % n % L 5 31.3 5 21.8 6 28.6 2 25.0 13 31.0 SL 7 43.8 4 17.4 6 28.6 1 12.5 6 14.3 SBV 1 6.2 1 4.3 3 14.2 1 12.5 9 21.4 N 3 18.7 13 56.5 6 28.6 4 50.0 14 33.3 Total 16 23 21 8 42 Combinations of viability classes L/L / / / / 1 10.0 / / 2 12.5 L/SL 2 28.6 2 18.2 2 20.0 / / 2 12.5 L/SBV / / / / / / / / 2 12.5 L/N 2 28.6 3 27.3 3 20.0 1 50.0 2 12.5 SL/SL 1 14.3 / / / / / / / / SL/SBV 1 14.3 / / 1 10.0 / / / / SL/N 1 14.3 1 9.1 2 20.0 1 50.0 3 18.8 SBV/SBV / / / / 1 10.0 / / 2 12.5 SBV/N / / 1 9.1 / / / / 1 6.2 N/N / / 4 36.4 1 10.0 / / 2 12.5 Total 7 11 10 2 16 In the year 2003, the distributions of the different viability classes in homozygotes for O-chromosomes in O-inversion karyotypes at different temperatures were similar (χ2= 8.29, DF = 6, p > 0.05). The grouping of some viability classes showed that the class of 'normals' was more frequent at 19°C and less frequent than expected at 12°C (χ2= 5.67, DF= 1, p < 0.05). Between 2003 cold stress (12°C) and 2005 mild cold stress (14–15°C) we observed a significantly different distribution for the semilethal (χ2= 5.72, DF= 1, p < 0.05) and quasinormal viability classes (SBV+N) (χ2=4.24, DF= 1, p < 0.05). The subvital class also showed a distinct distribution for these two years at nearly optimal temperature (19°C) in 2003 and mild cold stress (14–15°C) in 2005 (χ2= 3.23, DF= 1, p < 0.10). A similar result was obtained for the normal viability class (χ2= 3.19, DF=1, p < 0.10). The mean viability for the distinct O-inversion karyotypic combinations at different temperatures in the Avala Mountain population is presented in Table 4. The 2003 sample showed a higher mean viability of Ost/Ost O-inversion karyotypes at 19°C than at 12°C (F=7.54, DF=4, p < 0.10). In this sample, a significant difference in mean viability was found for total O-inversion karyotypes, which showed greater survival at 19°C than at 12°C (F=5.11, DF= 18, p < 0.05). However, no significant difference in mean viability was found for all the O-inversion karyotypes taken together at 19°C and 25°C (F = 2.61, DF = 20, p>0.05). In the 2005 sample (between 14°C and 15°C) a higher mean viability was detected for the Ost/O3+4 than for the O3+4/O3+4 karyotype (F = 6.30, DF = 5, p<0.10). Table 4. Mean viability values for different O-inversion karyotypic combinations observed during September 2003, 2004 and 2005 at different temperatures from Avala Mountain population. September 2003 September 2004 September 2005 O-karyotypes n Mean±SE 12°C n Mean±SE 19°C n Mean±SE 25°C n Mean±SE 19°C n Mean±SE 25°C n Mean±SE 14–15°C Ost/Ost 3 16.7±4.4 3 31.3±1.3 6 20.0±7.6 1 14.0±0.0 1 24.0±0.0 1 17.0±0.0 Ost/O3+4 2 28.5±4.5 2 31.0±2.0 2 19.5±6.5 / / / / 5 29.0±2.7 Ost/O3+4+1 1 18.0±0.0 1 25.0±0.0 1 34.0±0.0 / / / / 1 37.0±0.0 Ost/O3+4+6 1 33.0±0.0 1 35.0±0.0 1 39.0±0.0 / / / / 1 26.0±0.0 Ost/O3+4+22 / / / / / / / / / / 3 22.3±3.4 O6/O3+4 / / / / / / / / / / 2 15.5±10.9 O6/O3+4+6 / / / / / / 1 34.0±0.0 1 4.0±0.0 / / O3+4/O3+4 1 31.0±0.0 1 31.0±0.0 1 35.0±0.0 / / / / 2 21.0±0.0 O3+4/O3+4+1 / / 2 31.0±2.0 2 25.5 ±0.5 1 29.0±0.0 1 11.0±0.0 2 26.0±7.0 O3+4/O3+4+5 1 29.0±0.0 1 29.0±0.0 1 28.0±0.0 / / / / / / O3+4/O3+4+22 / / / / / / / / / / 2 22.0± 12.0 O3+4+1/O3+4+5 / / / / / / / / / / 1 24.0±0.0 The mean viability of homozygotes for O-chromosome arrangements in the samples taken in 2003, 2004 and 2005 over a range of temperatures was also studied. In the 2003 sample, the O3+4+1 chromosomal arrangement (in the case with lethal genes) showed a significantly higher mean viability at 19°C than at 25°C (F = 7.88, DF=4, p < 0.05). Furthermore, this arrangement had a significantly higher mean viability than Ost at 19°C (F = 6.90, DF=11, p < 0.05). The same results were observed when the samples from 2003 and 2004 at 19°C were pooled (F= 10.89, DF= 14, p < 0.01). In the 2005 sample (at 14–15°C), no significant differences in mean viability between the different O-arrangements were found. Finally, for the 2003, 2004 and 2005 samples, the mean viability of groups of O-inversion homo- and heterokaryotypes on the basis of frequency at different temperatures was analyzed. A significant difference between mean viability of O-inversion homo- and heterokaryotypes of moderate-high frequency was found only in the 2005 sample (F = 5.7, DF=11, p < 0.05). A comparison between the mean viabilities of total O-inversion homo- and heterokaryotypes independent of their frequency did not show any significant differences at any temperature. For the 2003 sample, a significant difference in mean viability was found for all O-inversion karyotypes (homo- and heterokaryotypes), which showed greater survival at 19°C than at 12°C. This is probably a consequence of the trend for mean viability differences for O-inversion homokaryotypes (generally of higher frequency) (F = 5.19, DF = 6, p < 0.10), but not for O-inversion heterokaryotypes, at these temperatures (F= 1.16, DF= 10, p>0.05). DISCUSSION Changes in the chromosomal polymorphism composition The high abundance of O-chromosomal arrangements found in Avala Mountain flies has also been reported in other populations from southeastern Europe (Zivanovic et al. 2002; Zivanovic and Marinkovic 2003; Zivanovic 2007) as well as in populations from central Europe (Fontdevila et al. 1983; Mestres et al. 1994; Orengo and Prevosti 1996; Rodríguez-Trellesetal. 1996; Rodríguez-Trelles and Rodríguez 1998; Solé et al. 2002; Balanyá et al. 2004). We also detected low IFR values, which are typical of many natural Drosophila subobscura populations in central Europe (Krimbas 1992, 1993). The frequencies of all 'cold-adapted' standard autosomes (Jst, Ust, Est, Ost) in this study, as well as in most other D. subobscura populations from this part of the Balkan region (five of six populations studied) are much lower (especially for Ust and Jst) (Zivanovic et al. 2002; Zivanovic 2007) than their geographical distribution in the Palearctic region during the 1960s, 1970s and 1980s (Krimbas 1992, 1993). In contrast, the frequency of several 'warm-adapted' non-standard chromosomal arrangements (J1, U1+2, U1+2+6, E1+2+9, O3+4) increased over the same period of time. These results are in accordance with the effects presented above, observed in the European region, of a move towards southern characteristics (a decrease in the frequency of 'cold-adapted' standard and an increase in the frequency of 'warm-adapted' non-standard chromosomal arrangements). A decrease in the frequency of standard 'cold-adapted' as well as an increase in the frequency of some non-standard 'warm-adapted' chromosomal arrangements has been described in recent decades in European D. subobscura populations (De Frutos and Prevosti 1984; Gosteli 1990; Orengo and Prevosti 1996; Rodríguez-Trelles et al. 1996; Solé et al. 2002; Balanyá et al. 2004) and in American colonizing populations (Balanyá et al. 2006). These results were attributed to an adaptation of chromosomal inversion polymorphism to a warmer climate in both Europe and America. For instance, the European heat wave of summer 2003 provided the first evidence of climatic ch