Add to Wishlist
Add to Wishlist


QUANTIFYING THE PREDICTABILITY OF EVOLUTION AT THE GENOMIC LEVEL is a well-researched Life Sciences Thesis/Dissertation topic, it is to be used as a guide or framework for your Academic Research.


Repeatable phenotypic evolution includes parallel and convergent evolution in independent populations in response to similar environmental challenges and implies natural selection. These repeated genetic changes suggest that predictable genetic changes can be identified.

However, the extent of predictability of evolution and the different circumstances in which evolution is more or less predictable remain unclear. This dissertation attempts to identify and quantify the degree to which evolution is predictable and studies different mechanisms that contribute to the evolution of Lycaeides butterflies.

I evaluate predictability in various contexts by testing for overlap in genomic loci associated with an evolving trait or associated with a specific evolutionary process. These contexts include comparing natural populations on a geographical and a temporal scale, comparing natural and laboratory populations, and comparing locations across the genome. In chapter 2, I investigated whether historical admixture can predict patterns of introgression (gene flow between species) in a contemporary hybrid zone using Lycaeides butterflies.

Here, I first show that both ancient and contemporary hybrid zones experience consistent selection which affects patterns of introgression and genomic composition of hybrids in a similar manner. Therefore, I can predict evolutionary patterns in one hybrid zone from another. In chapter 3, I assessed the predictability of genomic changes underlying a recent host plant shift in Lycaeides melissa butterflies.

Here, I show genomic changes accompanying this host shift are somewhat predictable depending on the contextual comparisons. Having studied the genomic basis of evolution in the previous two chapters, I address another novel mechanism underlying host plant adaptation in these butterflies. In chapter 4, I assess the sources of variation in the gut microbial community of Lycaeides melissa caterpillars.

Here, I show that caterpillar gut microbial communities vary over time and differ between frags and whole caterpillar samples. Diet (host plant) and butterfly population have limited effects on microbial communities. Collectively, these results demonstrate that I can use different contexts to study the predictability of evolution. However, the degree of predictability varies across different contextual approaches. Quantifying the extent to which evolution is predictable can be crucial in understanding the causes and consequences of evolutionary predictability.


Repeated phenotypic evolution constitutes the parallel or convergent evolution of traits in independent populations in response to similar environmental pressures [3]. Accumulating data now shows that phenotypic and genetic convergence can occur across several taxa [2, 4, 7, 18, 25, 26]. Evolutionary processes such as mutation and random drift do not cause similar evolutionary shifts repeatedly in response to environmental changes.

Therefore, repeated use of the same underlying genes during parallel and convergent phenotypic evolution can be indicative of constraints on genetic pathways and imply natural selection [17]. Understanding these constraints and their effects on phenotypic evolution can provide a possibility to predict genetic evolution.

Therefore, instances of repeated phenotypic evolution can provide an opportunity to identify and measure the predictable genetic changes underlying adaptive evolution and speciation.

Even though several instances of parallel and convergent evolution have been recorded, the stochastic and contingent nature of evolution has been a classic topic of discussion in biology and has been presented with contrasting views [16, 21, 22]. On one end some instances suggest the unpredictable nature of evolution [1, 15], on the other end the repeated evolution of specific traits and genetic changes in response to similar environmental pressures emphasizes that evolution can be predictable [20].

However, multiple selective agents and genetic background (such as standing genetic variation, large effect sizes, higher mutation rates, and linkage or epistatic relationships) can affect the probability of repeated use of the same genes in natural populations [23]. Therefore, it can be said that the predictability of evolution is not discrete and instead can lie along a Quantitative continuum [1].

Along these lines, quantification of the extent of predictability at different levels (such as genotypes, phenotypes, genomes, and mutations) and different scales (geographic, temporal, and genomic scales and comparisons within these scales) can help better understand the predictability of evolution.

By using different contexts to quantify the predictability of genetic changes, we can obtain a deeper understanding of how natural populations will cope with similar environmental challenges due to climate changes or human-mediated habitat changes. In addition, this can help us dissect the various underlying mechanisms which drive adaptive evolution and speciation.

The overarching goal of this dissertation is to better understand the predictability of evolution and the mechanisms driving adaptation to novel environments by using Lycaeides butterflies. These butterflies provide two interesting avenues to test for predictable evolution: the first is novel host plant colonization, and the second is the existence of an ancient and contemporary hybrid zone in a restricted spatial scale wherein the ancient hybrid zone spans a wide geographic range and the
contemporary hybrid zone inhabits a very small space.

Lycaeides melissa occurs throughout western North America and utilizes several species of legumes as their native hosts across their geographic range in North Western United States. Since the introduction of Medicago sativa (alfalfa) in their host range, some populations have started to colonize this novel host [19]. Along these lines, there is evidence suggesting that populations have persisted and adapted to alfalfa even though it is a poor host compared to the native legumes [5, 6, 24].

This case of novel host colonization provides an exciting opportunity to understand the predictability of genome-wide evolutionary changes associated with a host-plant shift in these butterflies. I use this case to quantify predictability and use several contexts for studying predictable genomic changes underlying a life-history trait. In addition, I also use this background of host-plant shift to understand the role of the larval gut microbial community in host use in L. melissa. Another aspect of Lycaeides biology, which makes it an ideal system to address my research goals in the occurrence of hybridization in several species of Lycaeides.

Several species of Lycaeides hybridize in specific geographic regions and have even formed hybrid lineages. Lycaeides ideas and Lycaeides, melissa is two of the 5 nominal species of Lycaeides butterflies that occur in North America [10, 13]. Along these lines, there is evidence that Lycaeides come into secondary contact in various regions across their geographic range and have hybridized to produce. three additional admixed lineages [8, 9, 11, 12]. I use this case of hybridization as another novel context to study predictability.

Using this background, I first identified and quantified the degree of evolutionary predictability in the following contexts: 1) different type of comparisons (among natural populations vs. between natural and experimental populations 2) different scales of comparisons (on a large geographic scale vs. pair of populations in close proximity; on a large temporal scale) and lastly, 3) comparison of different genomic regions (autosomes vs. sex chromosomes).

Second, I was broadly interested in understanding adaptive evolution in these butterflies by studying different mechanisms that drive adaptation and speciation. Host-plant use in herbivorous insects is an interesting case in which to understand adaptation in response to contemporary habitat changes.

Therefore, I used genomic and microbiome approaches to understand how L. melissa butterflies adapt to novel host plant alfalfa. Hybrid zones can also be used to understand adaptive evolution since introgression can provide novel genetic variation to adapt to environmental changes and can be useful to identify genetic barriers to gene flow [14]. Therefore, I use Lycaeides hybrid zones and a genomics approach to identify genetically differentiated regions in hybrid populations experiencing variable environmental pressures.

Together these approaches combined with genomics analyses help me better understand the evolution of Lycaeides butterflies over a geographic scale wherein I compare natural populations that vary in their geographic distribution (widespread vs. small distribution) and on a temporal scale (ancient vs. contemporary hybrid zone).

In Chapter 2, I approach evolutionary predictability by quantifying the predictability of genomic changes in the context of temporal comparisons in natural hybrid zones. Studies of replicate hybrid zones have found evidence of similar patterns of introgression across transects, but these studies generally focus on hybrid zones of a similar age. Whether there is consistency in patterns of introgression over time (at different stages of hybrid zone formation) is less clear.

In this chapter, I use relatively old admixed populations of Lycaeides melissa and Lycaeides ideas butterflies (admixture occurred about 14,000 YBP) and populations from a recent, active hybrid zone (hybridization started around 200 years ago and is ongoing) to ask if evolutionary patterns in old admixed populations can predict evolutionary dynamics in the current hybrid zone. Here, I asked two questions.

First, how well can I predict genomic regions which are most resistant to gene flow in recent active hybrid from patterns of ancestry in the admixed populations? Second, can I identify the processes which drive repeated patterns of introgression? I used genomic data analyses and genome annotation to first delineate candidate genomic regions which show excess local ancestry in ancient hybrids and genomic regions which show variable patterns of introgression in contemporary hybrids. By identifying these regions, I could pinpoint specific locations of the genome and their functional properties to better understand what traits underlie reproductive isolation in Lycaiedes hybrids.

I first found that several regions of the genome show excess ancestry in ancient hybrids and several regions restrict introgression in contemporary hybrids. These regions were spread across the autosomes and sex chromosomes. I then saw that similar regions of the genome experience restricted introversion across ancient and contemporary hybrid zones. Second, the level of consistency in the overlap of
genomic regions is quite high between the two-hybrid zones and this indicates that natural selection is the deterministic force driving these patterns of introgression across time.

These results highlight that quantification of the degree of predictability is possible over a large temporal scale and can be variable in different regions of the genome. In chapter 3, I measured the predictability of genome-wide evolutionary changes associated with a recent host shift in Lycaeides melissa. There are various contexts that can be used to study repeatable genomic changes underlying a phenotype.

In addition, quantification of the degree of predictability can be crucial in drawing conclusions about the processes driving repeatable patterns
of evolution. In this chapter, I used two different contextual approaches to quantify the extent. of predictability of patterns of evolutionary change in nature. First, I identified genomic regions. most associated with host-use in L. melissa and tested if these regions are enriched for specific functional properties.

Second, I compared instances of repeated evolution of host shifts across different populations of L. melissa across a geographic scale, and, third, I used SNP × performance associations in a laboratory experiment to predict evolutionary changes underlying host use in nature. I used genomic analyses and delineated genomic regions associated with host use in several L. melissa populations.

There were several regions which were associated with host use and these were distributed across autosomes and sex chromosome. I then found that there is evidence of parallel genomic changes underlying host use among natural populations spread across a wide geographic range.

There is a significant increase in predictability when I compare populations in close proximity and associated with the same host plant. Additionally, I could partially predict genomic regions associated with host use in nature from SNP × performance associations in a laboratory experiment.

However, in both these cases, I could not predict the direction of allele frequency changes in nature from those in the performance experiment. These results highlight how the degree of predictability can be variable in different contexts and quantifying predictability can indicate if stochastic or deterministic processes are driving genomic changes underlying adaptive evolution.

While addressing the main objectives of my research, I was also curious about the role of different mechanisms in adaptation to a different environments. I used host-plant adaptation in L. melissa as a case to understand adaptive evolution through different mechanisms. In chapter 3, I. address adaptation to a novel host plants by understanding the genomic basis of novel host plant use.

In chapter 4, I tried to understand host plant adaptation in these butterflies by dissecting the role of the gut microbiome in host use by assessing the sources of variation in the gut microbial community of Laclede melissa caterpillars. Host plant use in herbivorous insects is a complex life-history trait that is affected by several aspects of the biology of the organism.

Insect gut microbiome can facilitate or constrain host plant use. In this chapter, I ask two questions. First, can different aspects of insects and host plant biology affect Lycaeides melissa caterpillar gut microbiome? Second, does go.

Does caterpillar gut microbiome community interact with caterpillar performance? I use caterpillar rearing experiments and 16s rRNA microbiome sequencing of the host plant, caterpillar frass, and whole body to address these questions. I first find that caterpillar age and sample type (frass or whole-body) causes variation in gut microbial communities. However, diet (host plant) and population have a limited effect on the gut microbiome.

Second, I found that there is no association of caterpillar gut microbial communities with caterpillar performance. Our results provide general insights into the role of the gut microbiome in host plant use in Lepidoptera. Finally, in chapter 5, I summarize the findings of the previous three chapters and present a conclusion concerning the specific questions addressed in this dissertation.

Additional information


Thesis/Dissertation topic


Life Sciences

No of Chapters







There are no reviews yet.

Only logged in customers who have purchased this product may leave a review.