Insights into the Mechanistic Underpinnings for the Evolution of Population Stability in Drosophila, Based on Modeling of Laboratory Populations

Tung, Sudipta (2012) Insights into the Mechanistic Underpinnings for the Evolution of Population Stability in Drosophila, Based on Modeling of Laboratory Populations. Masters thesis, Indian Institute of Science Education and Research Kolkata.

PDF (BS-MS Dissertation of 07 MS 54) MS_thesis_Sudipta_Tung.pdf - Submitted Version Restricted to Repository staff only Download (2MB)

Abstract

In this study, I have proposed a model to study how different life-history features interact to determine the population dynamics of D. melanogaster. Although less complex (fewer parameters) and hence, easier to implement and manipulate than some of the previous studies (Mueller, 1988; Zeineddin and Jansen), this model was able to qualitatively and quantitatively capture many trends of the data from two independent long-term experiments. For example, the model generated time series in all the nutritional regimes closely approximate experimental results of a 49-generation long time-series experiment. Moreover, the ln(Nt+1/Nt) vs. Nt plot in all the four regimes (LL, LH, HH, HL) are similar in both simulation and experimental results. The curvilinear nature of the log return maps in LH and HH regimes are captured using this model which is not possible using widely used Ricker-like models. Box plot of the simulation data matches closely with the experimental ones suggesting good correspondence with respect to several distributional properties like mean, median and other percentiles. Discriminant function analysis shows clear difference in the dynamics in the four nutritional regimes with respect to the five carefully chosen statistical probes as suggested by Kendall (1999). The simulation results also group separately and overlap with corresponding experimental ones supporting well accordance between them. With only few biologically relevant changes in parameter values, a completely independent 36-generation experiment is simulated using this model. The simulation data mimics experimental results in terms of both distribution (box plot) and dynamics (FI plot) properties. This validates the model and in the process, the mechanistic reason behind the evolution of constancy stability in a previous study (Dey et al 2008)vi has been investigated which was unknown before. It is found that in the selected flies, FEJs, the reduction in critical minimum size requirement and basal fecundity of the females, because of ≈125 generation of selection for faster larval development and early reproduction, are primarily responsible for the enhanced constancy stability in those flies compared to the controls. In the later part, the values of the four life-history parameters used in this model are relaxed over a range to explore the model response in order to obtain some better insights regarding the system, which are otherwise difficult to obtain experimentally. The model predicts that increase in hatchability in a population with food limiting environment can lead to higher fluctuation of population size as well as decreased persistence stability. Further investigation shows that as hatchability increases, larval competition intensifies and pre-adult mortality increases. In this scenario, per-female fecundity increases as adult density decreases which finally leads to an increase in fluctuation index and extinction probability. In accordance with the results of the 36-generation experiment, increasing minimum critical size requirement for successful pupation increases the fluctuations in population size as well as the probability of extinction. Increasing basal female fecundity, γ is expected to increase population fluctuation and extinction probability. That has also been reflected in simulation outcomes. But it has been found that in extreme values of γ, FI reduces. It is suspected that in such extreme values per-female fecundity increases so much that irrespective of the parental number, the number of offspring generated is always so high that only few adults survive stochastically which also generate plenty of eggs in the next generation. Thus following only the adult number may seem to has less fluctuation as indicated by lower FI value in such scenarios but looking also at the egg dynamics reveals the underlying disparity between adult and egg dynamics.Decreasing δ is found to have similar effect on constancy and persistent stability as that of increasing γ. It is worth mentioning here that as providing yeast to the adults increases female fecundity even in low adult densities, in the present model the parameter γ is altered to capture the scenario of different qualities of adult nutrition. However, Mueller et al. (1994) modeled this scenario by reducing a parameter similar to δ, the adult density sensitivity to fecundity; whereas Jansen et al. (2005) disregarded δ in their model and put emphasis on γ like parameter. Our study indicates that both these parameters are analogous to each other, and any one should suffice to model the behaviour of the population.

Actions (login required)

View Item