Know Your Enemy : Genomes of Biological Control Agents

• Main Author: Ferguson, Kim B
• Format: Dissertation
• Language: English
• Published: ProQuest Dissertations & Theses 01.01.2020
• Subjects: 19th century; 20th century; Agriculture; Agronomy; Animal behavior; Animal sciences; Annotations; Arthropods; Behavioral Sciences; Bioengineering; Biology; Collaboration; Crops; Ecology; Ecosystem biology; Engineering; Entomology; Evolution; Genes; Genetic engineering; Genetics; Genomes; Genomics; Greenhouses; Horticulture; Invertebrates; Native species; Organic chemistry; Pesticides; Plant resistance; Population; Researchers; Toads
• ISBN: 9798516044878

Biological control is the use of an organism, the biological control agent (BCA), to control the population of another organism, the pest. Biological control is used within a variety of contexts, such as in open-field agriculture such as maize, or in greenhouses for crops such as tomatoes or cucumbers. BCAs are judged not only by their efficiency and effectiveness in controlling pests, but also by their ease or ability to be reared in commercial settings, stored for release, or the likelihood of what are called non-target effects (undesired effects such as preying upon native insects that are not pests). While BCAs can be used isolated in several different agricultural practices, they have especially high potential as a key part of Integrated Pest Management (IPM). IPM is an ecosystem approach to agriculture, where the entire system in question is assessed and the main goal is more healthy crops and less pesticides. At face value, this sounds like a win-win situation as reducing pesticide use is necessary to reduce the negative impact of agriculture on the environment, and consumers (in general) are increasingly moving towards products with less pesticides or to organic products. Indeed, BCAs are often part of organically produced crops and products. However, the total uptake of IPM and other use cases of BCAs is low, partially due to the perceived unreliability of BCAs as compared to conventional pesticides. One way to improve BCAs is to use a genetics approach with next-generation sequencing and genomics. The study of genetics is essentially about evolution and inheritance, and its application on BCAs is fairly straightforward: are the traits that we are interested in, such as parasitism rate for parasitoid wasps or starvation resistance for predatory bugs, 1) heritable, and 2) able to be improved without deleterious (side-)effects? Is there enough genetic variation within a population to select for improvement of these traits? While these scenarios can be used for improving the BCAs themselves, the way that BCAs are monitored, stored, and assessed for non-target effects can also be improved using genetics and/or genomics. The EU-funded research training network Breeding Invertebrates for Next Generation BioControl (BINGO-ITN) contained several projects that together aimed to improve the production and performance of BCAs using genetic variation, including the use of genomic techniques. The project involved 24 senior researchers, 13 PhD projects, and 12 partners from academia, and industrial and non-profit organizations throughout the EU. This thesis contains the work of one of those projects. Furthermore, this thesis is framed as an anthology of works on the possibilities and potential of genomics and genomic resources of BCAs. The intention of this thesis was to generate genomic results and resources for five biological control agents, complete with context and suggestions for future directions. My approach was to focus on the direct application of each species, giving clear indications of the possible uses of these newly generated genomic resources, while trying to be as open as possible with my science. I also used the concept of the life cycle of a genome project as a guide to completing or at least advancing genome projects. In Chapter 2, we perform a systematic review where we delved into literature on genetic variation of BCAs. We focussed our search term for papers that mentioned genetic variation and/or heritability, as well as biological control and all possible permutations therein. While our search initially returned nearly 3,000 hits, this was quickly narrowed down to include only species known to be used in biological control, as well as to the purpose of our study question. In the end, 69 papers fitted the bill, though the majority of these papers did not mention quantifiable measures of genetic variation, such as heritability (h2 or H2 ) or evolvability (CVA). From this review, we were able to shed light on the traits currently being studied in relation to those potentially important for improving biological control. Additionally, we made a case for including quantifiable measures of genetic variation in associate research and more transparent reporting practices in general. The remaining research chapters, Chapters 3-7, are arranged in order of end product, as a genome was not generated for all five species. In Chapter 3, we were working with Amblyseius swirskii, a predatory mite from the Eastern Mediterranean which is used around the world in a large variety of greenhouse crops. However, its initial collection was a fairly small source population, and concerns about resiliency and field performance led us to investigate the genetic variation of the commercial population. To do so, we set up an inbred line from the commercial population using a mother-son mating scheme over ten generations, while collecting eight wild populations in Israel for comparison purposes. Using whole-genome nanopore sequencing, we obtained just 512 Mbp of clean, corrected reads from the inbred line. While this was not enough for genome construction, as was the original goal, it was more than enough for microsatellite mining. Using six microsatellite loci, DNA was pooled for microsatellite analysis, a cost-effective alternative to individual genotyping that proved effective for this study. Our findings indicate that the commercial population had reduced genetic variation and far more differentiated than its wild counterparts. Given these results, we recommended increasing the scope to more commercial populations, more consistent monitoring of commercial lines for genetic variation, and to consider introgressing new material into the commercial lines. We recommend that the latter is done first in a test population to see whether genetic variation can be increased without hampering the biological control performance. In Chapter 4 we present the linked-read genome sequencing and assembly approach for Bracon brevicornis, an ectoparasitoid wasp that is currently being investigated around the world for biological control. For sequencing and assembly strategy, we went with a linked-reads approach, where a small amount of input material binds with barcodes to aid in an assembly that has low coverage but high accuracy. In addition to B. brevicornis material, we used a well-studied and sequenced organism (Solanum lycopersicum, tomato) as “carrier DNA” in the library preparation step. The resulting genome was 123 Mbp in size. We chose to use a linked-read assembly as it requires far less input material and can provide “phased” genomes, where areas of heterozygosity can be visualised instead of being removed as in other methods. The difficulty in getting enough genetic material from these wasps is partially due to the complementary sex determination (CSD) system within B. brevicornis that complicates inbreeding as it results in sterile diploid males. In the end, our solution for dealing with a low amount of input DNA also allowed us to identify a genomic region that is likely linked to the CSD mechanism in this species. Along with this in-depth investigation into CSD, we performed a protein comparison between B. brevicornis and two other braconid wasps to highlight the possibilities for comparative genomics with this genome, as well as an assessment of the accuracy of our ab initio-only assembly. The genome of the widely used parasitoid wasp Trichogramma brassicae (Chapter 5) was achieved through a hybrid approach, where short and long-read sequencing technology was used. The homozygosity of our inbred line, S301, was likely caused by a Wolbachia infection. This bacterial endosymbiont is found in a variety of insect hosts, and in some cases it can lead to femaleonly populations, where unfertilized eggs become female wasps instead of the usual male wasps. Three different assemblers were used, and five potential assemblies were narrowed down to one that went on to receive ab initio-, homology-, and evidence-based annotation. The final assembly size is 235 Mbp distributed over 1,572 contigs and contains 16,905 genes. The whole-genome sequencing from Chapter 5 went on to generate microsatellites in Chapter 6 on sister species Trichogramma evanescens. The question at hand here was related to ongoing efforts of monitoring both BCAs and their wild counterparts. The dispersal modes of Trichogramma spp. are generally considered to be either through direct or wind-based dispersal, and as such its dispersal range is quite small. However, recent observations of phoresy (a form of biological “hitch-hiking” behaviour) would increase the dispersal range to that of any butterflies that wasps are hitching a ride on. With eight German wild-caught lines and two Dutch lines, we used a mixture of population genetics and population genomics to explore this question. Microsatellites, an unannotated genome of T. evanescens, and pooled sequencing were used, and our analyses indicate that the populations show slight isolation-by-distance and strong differentiation between lines and within collection sites, but no clear latitudinal cline. While more investigation is necessary, our results combined with several in-field observations of phoresy, suggest that the dispersal range of Trichogramma may be larger than previously thought. The final genome within this thesis (Chapter 7) belongs to Nesidiocoris tenuis, a predatory bug used in greenhouses throughout the Mediterranean Basin, and is the second linked-read genome within this thesis. A single N. tenuis female was the basis for this genome, along with potential bacterial contaminants. These tag-alongs were removed through two decontamination pipelines, one of which also identified putative regions of lateral gene transfer (LGT). The total genome assembly size is 355 Mbp. Post decontamination, we performed an ab initio-, homology-, and evidence-based assembly that yielded 24,688 genes, prompting a compar