DISCUSSION:

In the light of the results and the statistical analysis of correlation between fitness and recombination, our experiment did not find any statistically significant correlation between fitness and recombination in flies bearing mutagenized CyO chromosome on the wildtype background. It rather turned out that the correlation between fitness and recombination that we found was just an artifact of randomness, and did not reflect a true relation between fitness and recombination in the flies used in our experiment.

There can be several reasons for not finding any statistically significant correlation between fitness and recombination. First, there might have been a correlation between fitness and recombination, but we did not have enough statistical power to detect it. Second, there might have been some flaws in the experiment design that rendered it unable to record any correlation between fitness and recombination. Third, there might not be any correlation at all between fitness and meiotic recombination.

As far as the statistics are concerned, our sample size might not have been large enough to bracket a significant correlation between fitness and recombination. Moreover our choice of the statistical treatment might not have been the most appropriate.

Alternatively, the experimental design might have been defective or inefficient to detect any significant correlation between the two properties. In fact there were several problems that occurred during the experiment. First, the incubators we raised our flies in were not equipped to regulate the humidity within. Therefore we had to water the vials every other day to keep the culture medium from drying up. However, the practice of watering vials was not quite precise, and sometimes an excess of even a drop could make the conditions more hostile for the larvae than favorable. In addition, it was very hard to keep the moisture constant across the vials, because not all the vials lost moisture at the same rate, some were drier than the others, therefore adding the same amount of liquid to each one of them could still not keep the moisture constant. Therefore as an alternate to watering individual vials, we attempted to raise the humidity of the entire incubator by putting in more sources of moisture (troughs of water). This helped the weaker lines survive better than before, however it over moistened the vials with good lines, especially when the excrement from the emerging offspring clogged the stoppers, thereby preventing the moisture from escaping the vial. Moreover, since we collected flies for scoring, 15 days after mating, by then, the stoppers of the vials with high density of flies were completely clogged with excrement from the emerging flies, and they were very wet inside. The highly moist conditions within the vials caused a lot of flies to stick to the walls of the vials and die or be rendered unidentifiable.

Therefore we lost a considerable chunk of data in the form of unidentifiable and dead flies, in many vials. This loss would have been even more significant if some of the lost individuals were double recombinants, because the double recombinants, although much less frequent than single recombinants, hold a lot more weightage to them than single recombinants. Loss of one double recombinant individual equals the loss of two single recombinants. The issue of loosing data due to flies dying gets even more complicated if all the eight phenotypic classes from the three point cross are not equally viable, which was the case in our experiment. Thus we might say that not having observed any significant correlation between fitness and recombination was due to the loss of data in the form of dead or unidentifiable individuals. This could especially be true if the correlation was a weak one. However, there still exists a possibility that there was not any correlation at all between fitness and recombination.

Besides the problems in following the designed protocol, there might have been some problems with the assumptions and speculations on which the protocol itself was based. The protocol assumed that EMS-induced mutagenesis would generate enough variability among the CyO balancers (20% of the diploid genome) across the lines to mask any variations among the lines that were due to other chromosomes (80%) or due to the environmental variation. Nonetheless, the statistical analysis revealed no significant variability among the lines (660 offspring²). Moreover, the probability of 0.97 of observing the correlation (between fitness and recombination) we observed, or larger, means that the three vials in each line were not more similar (with respect to recombination and fitness) to one another than they were to the vials in the other lines. If the variability within the lines was a lot less than that among the lines, then we would never have gotten such a high probability (0.97) of getting the observed correlation just by chance. This leads to the conclusion that the EMS mutagenesis did not carve the range of variation among the CyO balancers across the lines, which could mask any variability due to the non-mutagenized genome as well as due to the environment. A variation this steep, if generated, could have had a significant effect on both the fitness as well as recombination.

An improved design employing the same approach might involve deriving the non-mutagenized background from a highly inbred isogenic stock rather than the lab population, this way the variability due to the non-mutagenized background would be minimum. The dosage and choice of the mutagen is also important so as to achieve a variation among the mutagenized chromosomes, across the different lines, that is steep enough to mask the effects of other sources of variation. Moreover keeping the environment strictly constant is also important so as to avoid the environmental ‘noise’ from affecting the data.

REFERENCES:

Abdullah, M. F. F., and R. H. Borts, 2001 Meiotic recombination frequencies are affected by nutritional states in Saccharomyces cerevisiae. Proc. Natl. Acad. Sci. USA 98: 14524–14529.

Agrawal A.F., L. Hadany, and S.P. Otto.  2005. The evolution of plastic recombination. Genetics, 171: 803-812.

Balyaev, D. K., and P. M. Borodin, 1982 The influence of stress on variation and its role in evolution. Biol. Zentbl. 100: 705–714.

Grell, R. F., 1978 A comparison of heat and interchromosomal effects on recombination and interference in Drosophila melanogaster. Genetics 89: 65–77.

Kovalchuck, I., O. Kovalchuck, V. Kalck, V. Boyko, J. Filkowski et al., 2003 Pathogen-induced systematic plant signal triggers DNA rearrangements. Nature 423: 760–762.

Lucht, J. M., B. Mauch-Mani, H.-Y. Steiner, J.-P. Metraux, J. Ryals et al., 2002 Pathogen stress increases somatic recombination frequency in Arabidopsis. Nat. Genet. 30: 311–314.

Neel, J. V., 1941 A relation between larval nutrition and the frequency of crossing over in the third chromosome of Drosophila melanogaster. Genetics 26: 506–516.

Plough, H. H., 1917 The effect of temperature on crossing over in Drosophila. J. Exp. Biol. 24: 147–209.

Plough, H. H., 1921 Further studies on the effect of temperature on crossing over. J. Exp. Biol. 32: 187–202.

Zuchenko, A. A., A. B. Korol, T. A. Gavrilenko and T. Y. Kibenko, 1986 Correlation between stability of the genotype and the change in its recombination characteristics under temperature influences. Sov. Genet. 22: 730–737.