THE TAXONOMIC AND FUNCTIONAL DIVERSITY OF URBAN GROUND ARTHROPODS WITH A SPECIAL FOCUS ON PHORIDS by Claire Paillard B.Sc., McGill, 2020 THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE IN NATURAL RESOURCES AND ENVIRONMENTAL STUDIES-BIOLOGY UNIVERSITY OF NORTHERN BRITISH COLUMBIA July 2024 © Claire Paillard, 2024 i ABSTRACT Urban environments can host diverse arthropod communities that provide critical ecosystem services. Yet cities are complex, heterogenous habitats with many different land management practices at small spatial scales, often with variable effects on arthropods. Arthropods should be considered during urban development, which often involves habitat modification that alters arthropod biodiversity and the services they provide, such as pollination and pest-control. In this study, I investigated how urban land use types impact the taxonomic and functional diversity of ground arthropods. I compared arthropod communities from industrial, greenbelt, and residential land use types across twelve sites (n=4 per land use type) in Prince George, a mid-sized British Columbia city, sampled in 2015. I also focused on phorids (Diptera: Phoridae) in chapter two, which represent one of the most taxonomically rich and abundant insect groups in urban spaces. I sampled phorids from 30 sites during the summer of 2022. For ground arthropods, neither functional nor taxonomic diversity differed significantly between land use types. Composition of communities, however, was distinct and urbanized land use types favoured herbivorous taxa. For the 2022 data, 99 operational taxonomic units were detected with DNA barcoding, with greenbelt, edge, and residential areas harbouring the most diverse and abundant fly communities. Overall, Prince George hosts species-rich, functionally diverse arthropod communities, even in its most urbanized land use types. This suggests that highly modified habitats can be managed to support high arthropod diversity and ecosystem function, especially in large grassy areas and ii should be included in urban conservation projects. However, not all taxa tolerate industrial sites well, including phorids, which are some of the most taxonomically and ecologically diverse arthropods. Therefor, other land use types must be preserved to ensure that high gamma diversity in maintained. This work provides a foundation for management and preservation of biodiversity and ecosystem services in similar urban habitats. iii ACKNOWLEDGEMENTS I would like to extend an enormous thank you to Dezene Huber for encouraging and guiding me throughout this project, Lisa Poirier, for providing a lab space, expertise, and invaluable support, Che Elkin, and Felix Sperling for being a part of my committee and helping advise and improve my work, and Alicja Muir for collecting and sorting the 2015 data. I would also like to thank the BC Ministry of Advanced Education and Skills Training, the Fonds de Recherche du Quebec: Nature et technologies, and the Natural Sciences and Engineering Research Council of Canada for funding this research. Lastly, I would like to thank the Prince George residents, and business owners who let me sample their properties, and without whom this research would not have been possible. iv Table of contents Abstract……………………………………………………………………………………….ii Acknowledgements…………………………………………………………………………iv List of tables………………………………………………………………………………..viii List of figures………………………………………………………………………………...x Chapter one: Introduction…………………………………………………………………..1 1.1 BACKGROUND…………………………………………………………………………1 1.2. URBAN ENTOMOLOGY………………………………………………………………1 1.3. EFFECTS OF URBANIZATION ON TAXONOMIC DIVERSITY…………………..2 1.4. URBANOPHILE INSECTS……………………………………………………………3 1.5. FUNCTIONAL DIVERSITY AND ECOSYSTEM SERVICES……………………...4 1.6. TAXONOMIC DIVERSITY IN DIFFERENT LAND USE TYPES…………………..5 1.7. FUNCTIONAL DIVERSITY IN DIFFERENT LAND USE TYPES…………………5 1.8. INSECT CONSERVATION……………………………………………………………6 1.9. PROJECT GOALS……………………………………………………………………..8 1.10. REFERENCES………………………………………………………………………..9 Chapter two: The taxonomic and functional diversity of ground arthropods in different land use types……………………………………………………………………………...14 2.1 BACKGROUND………………………………………………………………………..14 2.2 OBJECTIVES………………………………………………………………………….14 2.3. PREDICTIONS………………………………………………………………………..15 2.4. METHODS…………………………………………………………………………….16 2.4.1. Study Sites…………………………………………………………………..16 2.4.2. Geospatial Land Feature Data…………………...……………………….19 2.4.3. Sampling…………………………………………………………………….19 2.4.4. Statistical Analyses…………………………………………………………20 2.4.4.2. Baseline of species richness in Prince George……………….21 2.4.4.3. Comparison of land use types…………………………………..21 2.4.4.4. Community similarity……………………………………………..22 v 2.4.4.5. Habitat preference………………………………………………..23 2.4.4.6. Geospatial modelling…………………………………………….23 2.4.4.7. Functional diversity………………………………………………24 2.4.4.8. Guild prevalence………………………………………………….28 2.4.4.9. Host specialization……………………………………………….29 2.5. RESULTS……………………………………………………………………………...29 2.5.1 Land use type classification………………………………………………..29 2.5.2. Species baseline……………………………………………………………30 2.5.2.1. Data quality……………………………………………………….30 2.5.2.2. Richness estimates………………………………………………31 2.5.3. Richness between land use types………………………………………..33 2.5.4. Geospatial modelling……………………………………………………….35 2.5.5. Community similarity……………………………………………………….37 2.5.6. Habitat preference………………………………………………………….41 2.5.7. Functional diversity…………………………………………………………43 2.5.8. Guild prevalence……………………………………………………………45 2.5.9. Host specialization………………………………………………………….47 2.6. DISCUSSION…………………………………………………………………………49 2.7. REFERENCES………………………………………………………………………..53 Chapter three: The abundance and diversity of scuttle flies in Prince George……..59 3.1. INTRODUCTION……………………………………………………………………..59 3.2. OBJECTIVES…………………………………………………………………………62 3.3. METHODS…………………………………………………………………………….63 3.3.1. Study sites…………………………………………………………………..63 3.3.2. Sampling…………………………………………………………………….65 3.3.3. Sorting and DNA sequencing……………………………………………..66 3.3.4. mOTU count………………………………………………………………...67 3.3.5. Abundance modelling………………………………………………………68 3.3.6. mOTU richness by land use type…………………………………………71 vi 3.3.7. Identifications and BINs……………………………………………………71 3.3.8. Phorid records in British Columbia from GBIF and BOLD……………..72 3.4. RESULTS……………………………………………………………………………...74 3.4.1. Data quality………………………………………………………………….74 3.4.2. Richness estimates………………………………………………………...77 3.4.3. Fly abundance modelling………………………………………………….81 3.4.4. mOTU richness by land use type…………………………………………89 3.4.5. Provincial data from GBIF and BOLD databases……………………….93 3.5. Discussion……………………………………………………………………………..97 3.6. References…………………………………………………………………………..101 Chapter four: Summary………………………………………………………………….105 5.1. Appendix I……………………………………………………………………………106 5.2. Appendix II……………………………………………………………………………117 vii List of tables Table 2.1. Estimates of OTU richness for the regional species pool and each land use type, along with their standard error for the chao 2 index, jackknife 1, jackknife2, and bootstrap estimators…………………………………………………………………………………32 Table 2.2. Comparisons of richness values between land use types where upper estimates of richness include all taxa and lower estimates exclude broader taxa. (df = degrees of freedom, res. df. = residual degrees of freedom, p = p-value)………33 Table 2.3. The results of simple linear regressions where the total vegetated area (tot. veg.) and Pielou’s evenness vegetation type (veg. evenness) are used to predict species richness (SR)………………..………………………………………………………………………36 Table 2.4. The results of two simple linear regressions; species richness (SR) as a function of grass area, and species richness as a function of forest area, both at a ten-meter scale……………………………………………………………………………………….36 Table 2.5. Results of ANOSIM (ANO) and PERMANOVA (PERM) tests which compared the similarity of greenbelt, residential, and industrial communities…….40 Table 2.6. The results of three Kruskal Wallis tests that compared functional richness values of three different indices against land use type……………………………..….43 Table 2.7. The results of three simple linear regressions that model functional richness as a function of species richness…………………………………..………….43 Table 2.8. The results of Kruskal-Wallis and Dunn tests comparing the proportions of different guilds against land use types……………………………………………..……46 Table 2.9. The associations of guilds with land use types. Correlations were calculated using the rpb index from the indicspecies package in R (permutations = 10000)……………………………………………………………………………………….46 Table 2.10. The results of Kruskal-Wallis tests for the proportions of generalist, oligophagous, and specialist OTUs compared between land use types, as well as the means between land use types…………………………………………………..….48 Table 3.1. Seven sampling periods used as random intercepts in mixed models..…70 Table 3.2. Model equations for nine mixed effects models (Abun = Abundance). Land use type is a 4-way factorial term………………………………………….………….....70 Table 3.3. The number of flies counted and sorted for each land use type. The mean ratio of morphospecies estimated (MSE) per specimens barcode, and the mean ratio of morphospecies estimated to the number of mOTUs detected……..……………...77 viii Table 3.4. Estimates of mOTU richness for the regional species pool and each land use type, along with their ± one standard error for the Chao 2 index, Jackknife 1, Jackknife2, Bootstrap, Chao 1, and Abundance Coverage estimators.……………..80 Table 3.5. Indices ranking the performance of nine fly abundance models, built using the compare_performance function from the performance package in R...…………86 Table 3.6. The results of likelihood ratio tests for nested sets of models fitted using restricted maximum likelihood. (Degrees of freedom= d.f., p-value = p, Abun = Abundance)……...................................................................................................….87 Table 3.7. The variable estimates for the full model of mean temperature……...…..88 Table 3.8. Comparison of mOTU richness by land use type and for 2015 and 2022 at selected dates with Kruskal-Wallis and Dunn tests…………………………………….90 Table 3.9. The number of mOTUs from PG belonging to each genus………….……94 Table 5.1.1. Table of OTU guild affinities………………………………………………106 Table 5.1.2. OTU observations by site and land use type (Gb. = Greenbelt, Ind. = Industrial, Res. = Residential)…………………………………………………………..109 Table 5.1.3. The association of OTUs with different land use types………………...113 Table 5.1.4. Degree of host specialization in herbivorous taxa where 1 indicates feeding within a single genus of plants. 2 within a single family, and 3 within multiple families…………………………………………………………………………………….116 Table 5.2.1. BINs and species names from the BOLD identification engine that matched OTUs from Prince George……………………………………………………117 Table 5.2.2. Number of phorid flies found in each land use type (Greenbelt = green, Edge = orange, Residential = blue, Industrial = gray), by year (2015 = 15, 2022 = 22) and the results of habitat associations found by the indicspecies package function multipatt……………………………………………………………………………..…….120 ix List of figures Figure 2.1. Map of 2015 study sites on a satellite map of Prince George from Google Earth…………………………………………………………………………………………18 Figure 2.2. NMDS Bray-Curtis dissimilarity plots where points represent sites and ellipses represent the standard error of weighted centroids…………………………..30 Figure 2.3. The OTU richness at each site……………………………………………...31 Figure 2.4. The OTU accumulation curve of OTUs detected from the 8645 specimens across 12 sites and four sampling periods………………………………...32 Figure 2.5. Dot plots representing upper (left column), and lower (right column) estimates for species, genus, and family richness between land use types………...34 Figure 2.6. Scatterplots of species richness as a function of total vegetated area (A) and as a function of Pielou’s evenness (B) for different vegetation types within a 10 m radius………………………………………………………………………………..……35 Figure 2.7. Linear regression plots of upper OTU richness estimates as a function of grass area (A) and forest area (B) within ten meters of a site………………………...36 Figure 2.8. A Venn Diagram of the number of OTUs that are unique or shared between land use types (constructed from pooled data for each land use type)…...38 Figure 2.9. Bray Curtis ordination (dissimilarity = "bray", dimensions = 4, permutations = 10000, stress = 0.052) plots arthropod communities in NMDS space where points represent sites, and color indicates land use type (meta.mds function from vegan in R)……………………………………………………………………………39 Figure 2.10. Bar plot of the habitat preference for different OTUs, where bars represent OTU richness and different colors represent habitat preference as described in the literature…………………………………………………………………42 Figure 2.11. The three box and whiskers plots above compare the median functional richness values (FRic, FDpg, FMPD) of different land use types, while the regression plots below show functional richness as a function of OTU richness………………..44 Figure 2.12. The box and whiskers plots compare the proportion taxa bellowing to herbivores (top left), natural enemies (top right), and decomposers (bottom left) out of all potential taxa at each site between land use types……………………………...45 Figure 2.13. The proportions of different host specialization categories in different colors over the total number of herbivores collected for each site……………………47 Figure 2.14. Box and whiskers plot of mean host specialization by land use type, calculated using the CWM.type method, in dbFD function from the FD package in R……………………………………………………………………………………………..48 x Figure 3.1. Map of 2022 study sites. on a satellite map of Prince George from Google Earth. Point color indicates land use types…………………………………….64 Figure 3.2. Abundance of flies by site for all dates (red), selected dates (2022-06-19 to 2022-09-11) which excluded early and late collections so that sites were comparable in their sampling effort (blue)………………………………………………74 Figure 3.3. The number of all, female, and male morphospecies estimated by site, the number of specimens barcoded, and the number of mOTUs actually detected for greenbelt, residential, edge, and industrial sites land use types……………………..76 Figure 3.4. The accumulation curve of mOTUs detected from the 582 sequenced specimens across 40 sites, over two sampling years (2015 and 2022)…………….78 Figure 3.5. The accumulation curves of mOTUs for Prince George where the black line represents the mean mOTU richness by site, red the richness estimated by the Chao 1 index, and blue the richness estimated by the Abundance Coverage index from estimateR in vegan…………………………………………………………..………79 Figure 3.6. The accumulation curves of mOTUs detected in greenbelt (green), residential (blue), and industrial (gray) land use types, using the specaccum (method=rarefaction) package from vegan in R……………………………………..…80 Figure 3.7. Mean temperature (°C) by date for the summer of 2022………………...83 Figure 3.8. Mean fly trapping rate over the summer of 2022 by land use type……..84 Figure 3.9. Temperature variation (temperature standard deviation/number of temperature readings) by collection dates………………………………………………85 Figure 3.10. Mean fly trapping rate by mean temperature with curves representing the slope of the model with mean temperature and land use type included as an intercept…………………………………………………………………………………….88 Figure 3.11. Mean fly trapping rate by mean temperature with slopes from the full model. Each land use type has its own slope and intercepts…………………………89 Figure 3.12. The number of mOTU detected for each site where color represents land use type and shape represents year……………………………………………….91 Figure 3.13. Venn diagram of mOTUs unique to and shared by land use types using sequenced flies from 2022 and 2015……………………………………………………92 Figure 3.14. Map of 12414 phorid observations downloaded from BOLD on September 24th, 2022.This map was built using the BCmaps and sf package……..95 Figure 3.15. The number of species and barcode index numbers (BINs) recorded in GBIF, and the Barcode of Life public database from September 2023 for the entire province of British Columbia and BC’s Sub-Boreal Interior ecoprovince specifically…………………………………………………………………………………..96 xi Figure 3.16. Venn diagram of species unique to and shared between GBIF, BOLD, and the Prince George project……………………………………………………………97 xii Chapter one 1.1. Background Cities are expanding. Conservative estimates project that global urban land area will double within this century (Gao et al. 2020), with roughly 68% of the world’s population expected to inhabit cites by 2050 (United Nations 2018). But urban planners must balance the need to develop against the need to protect rapidly changing ecosystems. Arthropods are by far the most abundant (Bar-On et al. 2018), and diverse group of animals on the planet (Hébert 2023). Many reports identify urbanization as one of the main culprits driving declines in global arthropod abundance and diversity (Fenoglio et al. 2020; van Klink et al. 2020). As cities develop, they break up natural habitat, create novel environments, expose native species to invasive competitors, and warm the surrounding area through the urban heat island effect (Fenoglio et al. 2020; Fenoglio et al. 2021). This can affect arthropod communities and the ecosystem services they provide. 1.2. Urban entomology Despite their importance, urban arthropods have long been overlooked by the scientific community. Urban entomology is a term that was coined in 1884 but wasn’t popularized until the 1970s. Most urban entomologists researched pest-control (Rust et al. 2024), until the early 2000s when more literature on the biodiversity of insects in cities began to emerge (Brown 2018). 1 Most urban animal biodiversity work focuses birds and mammals, leaving arthropods largely understudied (Collins et al. 2021).Unlike many larger animals, insects breed quickly, disperse easily, and need little space to persist (New 2015), allowing them to adapt to and even flourish in cities (Diamond et al. 2023). This makes them useful for gauging the effects of different kinds of disturbance (Kotze et al. 2011; Lewthwaite et al. 2024).Their ubiquity and incredible biological diversity also make them excellent subjects for urbanization studies, as well as bioindicators of ecosystem health (Schowalter et al. 2018). The need to understand the ecological effects of urbanization has fuelled a rise in the number of urban wildlife publications in the last decade, including those dealing with arthropods (Collins et al. 2021). The growing body of research has diverged to include lesser-known taxa and gathered sufficient data for researchers to conduct meta-analyses for various groups. This helps us understand how arthropods respond to urbanization at a global scale (Collins et al. 2024). 1.3. Effects of urbanization on taxonomic diversity The effects of urbanization on arthropod communities can be complex and diverse. While urbanization has been linked to widespread arthropod decline (van Klink et al. 2020), in other cases urbanization has yielded positive or no effects on arthropod diversity (Bang and Faeth 2011; Piano et al. 2020).This may partially be explained by how different taxa, from the species to the order level, respond differently to urbanization based on their respective niches. Bees, butterflies, and beetles are the most commonly studied groups, and often respond negatively to 2 urbanization (Merckx and Van Dyck 2019; Fenoglio et al. 2020; Fenoglio et al. 2021), while fruit flies do not (Avondet et al. 2003). Experimental design also explains variation between studies. Traditionally, urbanization has been explored along urban-rural gradients, but defining and comparing urban areas between studies can prove challenging (Moll et al. 2019). An urban area typically involves a developed area within a population center, but can include everything from landfills and industrial parks, to recreation grounds and gardens (New 2015). Studies may only compare areas within a city center, or include suburban areas, managed parks, and natural areas (Beninde et al. 2015; Turrini and Knop 2015). Additionally, rural areas can host lower diversity than urban ones when managed intensively (Turrini and Knop 2015). Studies that survey only a small area or short time period may not prove sufficient to capture large-scale or long-term trends (Fenoglio et al. 2020). Likewise, at a local scale, urbanization may increase biodiversity, but homogenize it at a regional scale if species being introduced in different cities are the same (McKinney 2006). 1.4. Urbanophile insects While many insects avoid highly built-up, disturbed urban environments, some species have adapted to tolerate, or even specialize to human associated habitats (Florencio et al. 2015). Pioneer and edge species often thrive in cities because open city spaces can resemble early successional stages (McKinney 2006). Herbivores can also reach high population densities in these areas due to predator release (Kruess and Tscharntke 1994). Additionally, cities often host introduced or invasive species because human transport concentrates around heavily populated areas 3 (Gippet et al. 2019). With climate change, many species ranges will shift, leading to new introductions, especially in urban areas. In temperate climates, biodiversity losses may be offset by new insects moving polewards, but not in tropical climates where many species have reached their maximum temperature thresholds (Diamond et al. 2023). 1.5. Functional diversity and ecosystem services Functional diversity is a field of research that examines a species’ ecological role or traits, rather than their taxonomy (Petchey and Gaston 2006; Gu et al. 2016). This approach is becoming increasingly popular in the literature as functional traits can often shed more light on how organisms interact with their environment than their taxonomic names alone (Petchey et al. 2009). This is especially relevant for arthropods as they provide many ecosystem services such as pollination, decomposition, and nutrient cycling (Schowalter et al. 2018). Often urbanization is detrimental to these services (Civeira et al. 2020). For example, as development intensifies, it can undermine ecosystem services ants provide like soil production and aeration (Sanford et al. 2009). When habitat fragmentation negatively affects predators and parasitoids, it can limit bio-control services (Kruess and Tscharntke 1994). Arthropod populations also forge a critical link in the food web. Fewer arthropods in cities can result in poorer diets for birds like great tits (Sinkovics et al. 2021), and increasing urban arthropod abundance could greatly improve bird diversity (Planillo et al. 2021). 4 1.6. Taxonomic diversity in different land use types Different sections of a city are zoned and managed to fulfill different needs. These can include industrial, commercial, residential, park, and recreational areas among others. Consequently, cities develop as mosaics of different kinds of land use types or habitat patches. Two patches might border each other, but undergo vastly different management regimes (Pickett et al. 2017). Industrial and residential sites tend to be planted with non-native species, regularly mown, and are subject to pesticide and fertilizer treatments (Proske et al. 2022). Green spaces or other natural or semi-natural areas are often far less maintained. This can alter the resources available to arthropod communities, their structure, and the ecological benefits they provide (McIntyre et al. 2001). Sites within the same land use type may also vary in their physical characteristics. Differences within land use types can reveal much about what conditions underlie biodiversity trends (McIntyre et al. 2001). For instance, urban ecologists measure gradients in vegetation cover, structure (Beninde et al. 2015; Herrmann et al. 2023), and the area of impervious surfaces to explore differences in arthropod communities (Su et al. 2011; Geslin et al. 2016). 1.7. Functional diversity in different land use types High levels of biodiversity are often assumed to improve ecosystem services, but this relationship depends on how species vary in their functional traits (Leps et al. 2006) –– high species richness does not necessarily equate with high functional diversity (Sobral et al. 2014; Freitas et al. 2021). When beneficial guilds of 5 arthropods like predators and pollinators are negatively impacted by urbanization, it can compromise ecosystem functions (Civeira et al. 2020), such as biological control services (Kruess and Tscharntke 1994), and pollination (Liang et al. 2023). To protect biodiversity and ecosystem function, researchers must consider functional traits along with species diversity (Auber et al. 2022). Urbanization filters out some species from communities based on the suitability of their traits related to their habitat (Gathof et al. 2022). The urban filter often favours generalist species at low trophic levels (Fenoglio et al. 2020). Likewise, the prevalence of different functional traits can reveal much about the selection pressures a community faces (Gerlach et al. 2013). Urbanization tends to shorten food chains (Kruess and Tscharntke 1994; Fenoglio et al. 2020) because natural enemies rely on prey populations and often need larger habitat areas to persist (Cardoso et al. 2011). This can lead to a loss of biological control services (Korányi et al. 2022). Urbanization also tends to select for dietary generalists because if a food source is compromised, generalist species can exploit other resources (Vázquez and Simberloff 2002; Concepción et al. 2015). 1.8. Insect conservation Few people appreciate the conservation potential of cities (New 2015). For the last century, most urban entomology research has focused on how to best eliminate unwanted insects. Pest species such as cockroaches, bedbugs, and termites thrive in urban conditions, and spread with human trade. This fuels a growing and costly demand for pest-control services and research (Rust et al. 2024). Unfortunately, urbanization can also deepen the public’s aversion to insects, 6 including harmless species (Fukano and Soga 2021). When people rarely encounter insects, they learn little of native species (McKinney 2006), and tend to think of insects as only pests in their home, further undermining conservation efforts (Fukano and Soga 2021). However, insects are fundamental to many ecosystem processes (Lewthwaite et al. 2024). Since urban areas are one of the fastest growing habitat types (Menke et al. 2011), and many insects species are suffering steep population declines (Habel et al. 2019), working to improve urban insect diversity is one of the few ways to reconcile development and conservation interests. Likewise, some areas in cities can be biodiversity hotspots (Connor et al. 2002), unique novel ecosystems, and havens for vulnerable species (Hall et al. 2017). In the last 25 years, there has been a growing recognition of the need to protect insects and explore their diversity in cities (Brown 2018; Collins et al. 2024). This is reflected in the literature by the growing number of urban biodiversity studies (Collins et al. 2021), surveys of understudied urban taxa (Hartop et al. 2015), and proposals for ways to improve insect conservation (Samways et al. 2020; Brandl et al. 2022). The awareness of issues insects face has helped both scientists and the public reimagine cities from ecological deserts to potential refuges for insects amidst widespread biodiversity loss due to climate change and human habitat destruction (Diamond et al. 2023). However, much work remains to be done. Media reports of an “insect apocalypse” have sensationalized real invertebrate declines, to point of becoming counter productive. In truth, many arthropods remain undescribed, and often little is 7 known of the life history of discovered species (Saunders et al. 2020). Many of the most widely cited papers discuss the general effects of urbanization on insect communities (habitat suitability). Few focus on faunistics, insect biology, or ecological theory (Brown 2018). Likewise, many solutions have been proposed to help support insect diversity: everything from bee hotels to planting gardens of native host plants. However, conservation efforts often target narrow groups of insects, and if left untested, these strategies risk being ineffective, or even becoming ecological traps (Brown 2018; Diamond et al. 2023). Future research needs to consider more nuanced approaches that investigate mechanisms behinds how arthropods respond to different urban disturbances, as well as vet restoration approaches to ensure they’re effective (Samways et al. 2020; Diamond et al. 2023). It is also clear that conservation cannot done by scientists alone. Generally, insects are underappreciated by most people, and disliked by many (Fukano and Soga 2021). Their importance must be communicated to stakeholders, policy makers, and the public for conservation efforts to succeed. Promoting citizen science and public education can help people realize the value of insects, increase support for conservation, and improve well-being for both insects and humans (Samways et al. 2020). 1.9. Project goals This project seeks to determine how land use type impacts the taxonomic and functional diversity of arthropod communities, as well as inventory which species are present, as northern British Columbia is comparatively understudied. As phorids are one of the largest and most ecologically diverse families of insect, I focus further 8 analysis on this group. I aim to quantify their abundance and diversity in different land use types and assess how they respond to environmental variables like season and temperature. 1.10. REFERENCES Auber A, Waldock C, Maire A, Goberville E, Albouy C, Algar AC, McLean M, Brind’Amour A, Green AL, Tupper M, et al. 2022. A functional vulnerability framework for biodiversity conservation. Nat. Commun. 13(1):4774. https://doi.org/10.1038/s41467-022-32331-y. Avondet JL, Blair RB, Berg DJ, Ebbert MA. 2003. Drosophila (Diptera: Drosophilidae) response to changes in ecological parameters across an urban gradient. Environ. Entomol. 32(2):347–358. https://doi.org/10.1603/0046-225X-32.2.347. Bang C, Faeth SH. 2011. Variation in arthropod communities in response to urbanization: Seven years of arthropod monitoring in a desert city. Landsc. Urban Plann. 103(3):383–399. https://doi.org/10.1016/j.landurbplan.2011.08.013. Bar-On YM, Phillips R, Milo R. 2018. The biomass distribution on Earth. Proc. of the Natl. Acad. of Sci. U.S.A. 115(25):6506–6511. https://doi.org/10.1073/pnas.1711842115. Beninde J, Veith M, Hochkirch A. 2015. Biodiversity in cities needs space: A meta-analysis of factors determining intra-urban biodiversity variation. Ecol. Lett. 18(6):581–592. https://doi.org/10.1111/ele.12427. Brandl M, Hussain RI, Maas B, Rabl D, Pachinger B, Holzinger W, Krautzer B, Moser D, Frank T. 2022. Improving insect conservation values of agri-environment schemes through diversified seed mixtures. Biol. Conserv. 269. http://dx.doi.org/10.1016/j.biocon.2022.109530. Brown B. 2018. After “the call”: a review of urban insect ecology trends from 2000–2017. Zoosymposia. 12:4. https://doi.org/10.11646/zoosymposia.12.1.3. Cardoso P, Pekár S, Jocqué R, Coddington JA. 2011. Global patterns of guild composition and functional diversity of spiders. PLOS One. 6(6):e21710. https://doi.org/10.1371/journal.pone.0021710. Civeira G, Lado Liñares M, Vidal Vazquez E, Paz González A. 2020. Ecosystem services and economic assessment of land uses in urban and periurban areas. Environ. Manage. 65(3):355–368. https://doi.org/10.1007/s00267-020-01257-w. Collins CM (Tilly), Audusseau H, Hassall C, Keyghobadi N, Sinu PA, Saunders ME. 2024. Insect ecology and conservation in urban areas: An overview of knowledge and needs. Insect Conserv. and Diver. 17(2):169–181. https://doi.org/10.1111/icad.12733. Collins MK, Magle SB, Gallo T. 2021. Global trends in urban wildlife ecology and conservation. Biol. Conserv. 261:1–8. https://doi.org/10.1016/j.biocon.2021.109236. 9 Concepción ED, Moretti M, Altermatt F, Nobis MP, Obrist MK. 2015. Impacts of urbanisation on biodiversity: the role of species mobility, degree of specialisation and spatial scale. Oikos. 124(12):1571–1582. https://doi.org/10.1111/oik.02166. Connor EF, Hafernik J, Levy J, Lee Moore V, Rickman JK. 2002. Insect conservation in an urban biodiversity hotspot: The San Francisco Bay Area. J. Insect Conserv. 6(4):247–259. doi:10.1023/A:1024426727504. Diamond SE, Bellino G, Deme GG. 2023. Urban insect bioarks of the 21st century. Curr. Opin. Insect Sci. 57:1–7. https://doi.org/10.1016/j.cois.2023.101028. Fenoglio MS, Calviño A, González E, Salvo A, Videla M. 2021. Urbanisation drivers and underlying mechanisms of terrestrial insect diversity loss in cities. Ecol. Entomol. 46(4):757– 771. https://doi.org/10.1111/een.13041. Fenoglio MS, Rossetti MR, Videla M, Baselga A. 2020. Negative effects of urbanization on terrestrial arthropod communities: A meta-analysis. Baselga A, editor. Glob. Ecol. Biogeogr. 29(8):1412–1429. https://doi.org/10.1111/geb.13107. Florencio M, Lobo JM, Cardoso P, Almeida-Neto M, Borges PAV. 2015. The colonisation of exotic species does not have to trigger faunal homogenisation: Lessons from the assembly patterns of arthropods on oceanic islands. Chapman M (Gee) G, editor. PLOS One. 10(5):e0128276–e0128276. https://doi.org/10.1371/journal.pone.0128276. Freitas C, Brum FT, Cássia-Silva C, Maracahipes L, Carlucci MB, Collevatti RG, Bacon CD. 2021. Incongruent spatial distribution of taxonomic, phylogenetic, and functional diversity in neotropical cocosoid palms. Fron. For. and Glob. Change. https://doi.org/10.3389/ffgc.2021.739468 Fukano Y, Soga M. 2021. Why do so many modern people hate insects? The urbanization– disgust hypothesis. Sci. Total Environ. 777:146229. https://doi.org/10.1016/j.scitotenv.2021.146229. Gao J, O’Neill BC, Iowa State Univ. A IA (United States). 2020. Mapping global urban land for the 21st century with data-driven simulations and shared socioeconomic pathways. Nat. Commun. 11(1):2302–2302. https://doi.org/10.1038/s41467-020-15788-7. Gathof AK, Grossmann AJ, Herrmann J, Buchholz S. 2022. Who can pass the urban filter? A multi-taxon approach to disentangle pollinator trait–environmental relationships. Oecologia. 199(1):165–179. https://doi.org/10.1007/s00442-022-05174-z. Gerlach J, Samways M, Pryke J. 2013. Terrestrial invertebrates as bioindicators: An overview of available taxonomic groups. J. Insect Conserv. 17(4):831–850. doi:10.1007/s10841-013-9565-9. Geslin B, Le Féon V, Folschweiller M, Flacher F, Carmignac D, Motard E, Perret S, Dajoz I. 2016. The proportion of impervious surfaces at the landscape scale structures wild bee assemblages in a densely populated region. Ecol. Evol. 6(18):6599–6615. doi:10.1002/ece3.2374. Gippet JM, Liebhold AM, Fenn-Moltu G, Bertelsmeier C. 2019. Human-mediated dispersal in insects. Curr. Opin. in Insect Sci. 35:96–102. doi:10.1016/j.cois.2019.07.005. 10 Gu J, Zhou J, Wilson M, Jia K, Lv K, Xu Z. 2016. Species diversity and functional diversity of insects in Wuxijiang National Wetland Park, East China. Acta Ecol. Sin. 36(5):386–391. doi:10.1016/j.chnaes.2016.07.002. Habel JC, Samways MJ, Schmitt T. 2019. Mitigating the precipitous decline of terrestrial European insects: Requirements for a new strategy. Biodivers. Conserv. 28(6):1343–1360. doi:10.1007/s10531-019-01741-8. Hall DM, Camilo GR, Tonietto RK, Ollerton J, Ahrné K, Arduser M, Ascher JS, Baldock KCR, Fowler R, Frankie G, et al. 2017. The city as a refuge for insect pollinators. Conserv. Biol. 31(1):24–29. doi:10.1111/cobi.12840. Hartop EA, Brown BV, Disney RHL. 2015. Opportunity in our ignorance: urban biodiversity study reveals 30 new species and one new Nearctic record for Megaselia (Diptera: Phoridae) in Los Angeles (California, USA). Zootaxa. 3941(4):451–484. Hébert C. 2023. Forest Arthropod Diversity. In: D. Allison J, Paine TD, Slippers B, Wingfield MJ, editors. Forest Entomology and Pathology: Volume 1: Entomology. Cham: Springer International Publishing. p. 45–90. https://doi.org/10.1007/978-3-031-11553-0_3. Herrmann J, Buchholz S, Theodorou P. 2023. The degree of urbanisation reduces wild bee and butterfly diversity and alters the patterns of flower-visitation in urban dry grasslands. Sci. Rep. 13(1):2702. doi:10.1038/s41598-023-29275-8. van Klink R, Bowler DE, Gongalsky KB, Swengel AB, Gentile A, Chase JM. 2020. Metaanalysis reveals declines in terrestrial but increases in freshwater insect abundances. Science. 368(6489):417–420. doi:10.1126/science.aax9931. Korányi D, Egerer M, Rusch A, Szabó B, Batáry P. 2022. Urbanization hampers biological control of insect pests: A global meta-analysis. Sci. Total Environ. 834:155396. doi:10.1016/j.scitotenv.2022.155396. Kotze J, Venn S, Niemelä J, Spence J. 2011. Effects of urbanization on the ecology and evolution of arthropods. In: Niemelä J, Breuste JH, Elmqvist T, editors. Urban ecology, patterns, processes and applications. Oxford University Press, New York:159–166. Kruess A, Tscharntke T. 1994. Habitat fragmentation, species loss, and biological control. Science. 264(5165):1581–1584. doi:10.1126/science.264.5165.1581. Leps J, de Bello F, Lavorel S, Berman S. 2006. Quantifying and interpreting functional diversity of natural communities: practical considerations matter. Preslia. 78(4):481–501. Lewthwaite JMM, Baiotto TM, Brown BV, Cheung YY, Baker AJ, Lehnen C, McGlynn TP, Shirey V, Gonzalez L, Hartop E, et al. 2024. Drivers of arthropod biodiversity in an urban ecosystem. Sci. Rep. 14(1):390. doi:10.1038/s41598-023-50675-3. Liang H, He Y-D, Theodorou P, Yang C-F. 2023. The effects of urbanization on pollinators and pollination: A meta-analysis. Ecol. Lett. 26(9):1629–1642. doi:10.1111/ele.14277. McIntyre NE, Rango J, Fagan WF, Faeth SH. 2001. Ground arthropod community structure in a heterogeneous urban environment. Landsc. Urban Plann. 52(4):257–274. doi:10.1016/S0169-2046(00)00122-5. 11 McKinney ML. 2006. Urbanization as a major cause of biotic homogenization. Biol. Conserv. 127(3):247–260. doi:10.1016/j.biocon.2005.09.005. Meidinger D, Pojar J. 1991. Chapter 14: Sub-Boreal Spruce Zone. In: Ecosystems of British Columbia. Victoria, British Columbia: British Columbia Ministry of Forests. (Special Report Series). p. 210–221. http://www.for.gov.bc.ca/hfd/pubs/Docs/Srs/SRseries.htm. Menke SB, Guénard B, Sexton JO, Weiser MD, Dunn RR, Silverman J. 2011. Urban areas may serve as habitat and corridors for dry-adapted, heat tolerant species: An example from ants. Urban Ecosys. 14(2):135–163. doi:10.1007/s11252-010-0150-7. Merckx T, Van Dyck H. 2019. Urbanization-driven homogenization is more pronounced and happens at wider spatial scales in nocturnal and mobile flying insects. Glob. Ecol. Biogeogr. 28(10):1440–1455. doi:10.1111/geb.12969. Moll RJ, Cepek JD, Lorch PD, Dennis PM, Tans E, Robison T, Millspaugh JJ, Montgomery RA. 2019. What does urbanization actually mean? A framework for urban metrics in wildlife research. J. Appl. Ecol. 56(5):1289–1300. doi:10.1111/1365-2664.13358. New TR. 2015. Insect conservation and urban environments. New York;Cham, Switzerland; Springer. https://go.exlibris.link/bC3xPLmJ. Petchey O, O’Gorman E, Flynn D. 2009. A functional guide to functional diversity measures. In: Biodivers. Ecosyst. Funct. Hum. Wellbeing Ecol. Econ. Perspect. Vol. 85. Petchey OL, Gaston KJ. 2006. Functional diversity: back to basics and looking forward. Ecol. Lett. 9(6):741–758. doi:10.1111/j.1461-0248.2006.00924.x. Piano E, Souffreau C, Merckx T, Baardsen LF, Backeljau T, Bonte D, Brans KI, Cours M, Dahirel M, Debortoli N, et al. 2020. Urbanization drives cross-taxon declines in abundance and diversity at multiple spatial scales. Glob. Change Biol. 26(3):1196–1211. doi:10.1111/gcb.14934. Pickett STA, Cadenasso ML, Rosi-Marshall EJ, Belt KT, Groffman PM, Grove JM, Irwin EG, Kaushal SS, LaDeau SL, Nilon CH, et al. 2017. Dynamic heterogeneity: a framework to promote ecological integration and hypothesis generation in urban systems. Urban Ecosyst. 20(1):1–14. doi:10.1007/s11252-016-0574-9. Planillo A, Kramer-Schadt S, Buchholz S, Gras P, von der Lippe M, Radchuk V. 2021. Arthropod abundance modulates bird community responses to urbanization. Divers. Distrib. 27(1):34–49. doi:10.1111/ddi.13169. Proske A, Lokatis S, Rolff J. 2022. Impact of mowing frequency on arthropod abundance and diversity in urban habitats: A meta-analysis. Urban For. Urban Green. 76:127714. doi:10.1016/j.ufug.2022.127714. Rust MK, Lee C-Y, Bennett GW, Robinson WH. 2024. The emergence and sustainability of urban entomology. Annu. Rev. Entomol. 69(Volume 69, 2024):59–79. doi:https://doi.org/10.1146/annurev-ento-012423-110612. 12 Samways MJ, Barton PS, Birkhofer K, Chichorro F, Deacon C, Fartmann T, Fukushima CS, Gaigher R, Habel JC, Hallmann CA, et al. 2020. Solutions for humanity on how to conserve insects. Biol. Conserv. 242:108427. doi:10.1016/j.biocon.2020.108427. Sanford Monte P., Manley Patricia N., Murphy Dennis D. 2009. Effects of urban development on ant communities: Implications for ecosystem services and management. Conserv. Biol. 23(1):131–141. doi:10.1111/j.1523-1739.2008.01040.x. Saunders ME, Janes JK, O’Hanlon JC. 2020. Moving on from the insect apocalypse narrative. BioSci. 70(1):80–89. Schowalter TD, Noriega JA, Tscharntke T. 2018. Insect effects on ecosystem services— Introduction. Basic Appl. Ecol. 26:1–7. doi:10.1016/j.baae.2017.09.011. Sinkovics C, Seress G, Pipoly I, Vincze E, Liker A. 2021. Great tits feed their nestlings with more but smaller prey items and fewer caterpillars in cities than in forests. Sci. Rep. 11(1):24161–24161. doi:10.1038/s41598-021-03504-4. Sobral FL, Jardim L, Lemes P, Machado N, Loyola R, Cianciaruso MV. 2014. Spatial conservation priorities for top predators reveal mismatches among taxonomic, phylogenetic and functional diversity. Nat. Conservação. 12(2):150–155. doi:10.1016/j.ncon.2014.09.008. Su Z, Zhang R, Qiu J. 2011. Decline in the diversity of willow trunk-dwelling weevils (Coleoptera: Curculionoidea) as a result of urban expansion in Beijing, China. J. Insect Conserv. 15(3):367–377. doi:10.1007/s10841-010-9310-6. Turrini T, Knop E. 2015. A landscape ecology approach identifies important drivers of urban biodiversity. Glob. Change Biol. 21(4):1652–1667. doi:10.1111/gcb.12825. United Nations. 2018. United Nations Department of Economic and Social Affairs. 68% of the world population projected to live in urban areas by 2050, says UN. https://www.un.org/development/desa/en/news/population/2018-revision-of-worldurbanization-prospects.html. Vázquez DP, Simberloff D. 2002. Ecological specialization and susceptibility to disturbance: Conjectures and refutations. Am. Nat. 159(6):606–623. doi:10.1086/339991. 13 Chapter two: The taxonomic and functional diversity of ground arthropods in different land use types 2.1. Background As cities develop, managers must consider how to best meet human demands, but also preserve or improve ecosystem health and the benefits arthropods provide. In my research I investigated how urbanization affects the taxonomic and functional diversity of ground arthropods in Prince George, a midsized British Columbia (BC) city. Urban biodiversity studies typically take place in larger, southerly cities, leaving the insect fauna of northern BC largely understudied. The city of Prince George provides an opportunity to study insect biodiversity in the context of an urbanized area mostly surrounded by forests. 2.2. OBJECTIVES The main objectives for the first part of my thesis research were to establish a baseline of which ground-dwelling arthropod species were present in Prince George, explore their taxonomic diversity, and compare those trends with patterns of functional diversity across different land use types using pitfall trap-derived data from industrial, residential, and greenbelt areas. I aimed to answer four questions: i) How does land use type impact taxonomic diversity? ii) How do different local physical land features impact species richness? iii) How does land use type impact functional diversity? 14 iv) Do different land use types exert stronger selection pressure for certain traits in arthropod communities such as host specialization and habitat preference? 2.3. PREDICTIONS i) Greenbelt and residential sites should have the highest species richness as they have the more vegetated area within a 50 m radius compared to industrial sites. I expected residential sites to have higher species richness than industrial sites but similar or fewer species than greenbelt sites as they had less vegetation, but also very varied vegetation and additional food sources such as composts, trash, and gardens. ii) a) Species richness would be positively related to total vegetation and vegetation evenness (measured as the Pielou evenness index for trees, shrubs, grass and gardens at each site). Vegetated area is considered favourable habitat to arthropods, so increasing vegetation increases potential habitat for arthropod taxa (Beninde et al. 2015; Turrini and Knop 2015). Increasing habitat heterogeneity can also increase species richness (Cramer and Willig 2005), so increasing vegetation evenness (more balanced vegetation structure), should increase diversity. b) Industrial and greenbelt communities were most different in terms of their geospatial variables (high lawn and hard surface cover compared to high tree cover, respectively), so I also expected them to be very different in terms of community composition. 15 iii) Functional diversity would be lowest in industrial sites. Industrial sites lacked resources like woody debris, and gardens which supplied flowers, which could be important to decomposers (Tóth et al. 2021), and natural enemies respectively (Corcos et al. 2019). iv) a) There would be higher proportions of natural enemies at greenbelt sites compared to industrial sites as greenbelt sites had more vegetation cover and could potentially support more predator and parasitoid taxa by providing habitat for both herbivore prey and the natural enemies (Kruess and Tscharntke 1994). b) There would be more polyphagous taxa and habitat generalists/open habitat taxa at industrial sites. Urbanization typically selects for dietary and habitat generalists (Concepción et al. 2015; Magura et al. 2020), so I expected more heavily managed and developed land use types to support more generalist taxa. 2.4. METHODS 2.4.1. Study sites Prince George falls within the Sub-Boreal Spruce biogeoclimatic zone, which supports mostly mixed forests of gymnosperm (e.g., Picea spp., Pinus contorta, Pseudotsuga menziesii) and angiosperm (e.g. Populus spp., Betula papyrifera, Alnus spp.) trees and associated understory plants. It experiences warm, wet, short summers and long cold winters (Meidinger and Pojar 1991). The mean annual 16 temperature of the region has warmed by 1.3oC in the last century, and will likely rise another 2.2oC to 3.7oC in the next 60 years (Coady and Picketts 2012). Prince George supports a population of 77000 people (Statistics Canada 2023) spread over an area of 329 km2 (City of Prince George 2023). It encompasses several industrial zones including pulp mills, an oil refinery and chemical plant (Prince George Area Industrial Land Profile 2008), along with residential areas of mostly low-density housing (City of Prince George 2023). The city contains at least 6000 hectares of productive forest on municipal and crown land (Timberline Forest Inventory Consultants 2006), and 65% of the land is considered greenbelt when including parks together with natural forests (Coady and Picketts 2012). The study area for this project encompassed 50 km2 in the main urbanized portion of the city – much of the outlying area remains unurbanized – in three different land use types: industrial, residential, and greenbelt spaces. Industrial areas consisted of local businesses dominated by lawns and pavement. Commercial areas were included with industrial areas in this project as both contained lots of hard surfaces and lawns. Residential areas included private residences with lawns and gardens, while greenbelt areas were forested spaces within and around Prince George generally lacking substantial management or roadways. The residential sites in this study were confirmed to be pesticide free with homeowners, while industrial and greenbelt sites (within treeline) likely were as well. Industrial and greenbelt sites were selected by driving in Prince George and randomly choosing sites in four of the city’s industrial areas. Residential sites were selected by recruiting local 17 homeowners to the project and selecting residences that were situated in different parts of the city. Site R1 R2 R3 R4 G1 G2 G3 G4 I1 I2 I3 I4 Latitude 53.856950° 53.873243° 53.898214° 53.919023° 53.870774° 53.903715° 53.936776° 53.910720° 53.878739° 53.907891° 53.941743° 53.913714° Longitude -122.755560° -122.775723° -122.764266° -122.770138° -122.777743° -122.809661° -122.820802° -122.744167° -122.738973° -122.787999° -122.821441° -122.736775° Figure 2.1. Map of 2015 study sites on a satellite map of Prince George from Google Earth. Colored points indicate land use types. Sites (R = Residential, G = Greenbelt, I = Industrial) are shown in the table beside. 18 2.4.2. Geospatial land feature data Using 2014 orthophoto data of the sites provided by the City of Prince George, circles were drawn around each site with 50 m, 25 m, and 10 m radii. Physical features such as tree canopy, sidewalk, roads, gravel, buildings, lawn etc. were measured in pixels (~1m) within each circle. I combined these features into seven variables: the area of total vegetation, trees, shrubs, lawns, gardens, hard surfaces, and buildings for each site. Hard surfaces included gravel and pavement. Different types of vegetation included both planted and naturally occurring vegetation. 2.4.3. Sampling To survey Prince George, five pitfall traps were set in four-meter transects at twelve sites in industrial, residential, and greenbelt land use types (n = 4 sites for each land use type) with traps spaced one meter from each other. The traps contained 50% propylene glycol and were collected every week from July to late August 2015. Specimens were stored in 95% ethanol and initially processed by students in Dezene Huber and Lisa Poirier’s labs using guides such as Ubick et al.’s 2017 Spiders of North America: An Identification Manual (2 nd ed.), Roth’s 1993 Spider Genera of North America (3rd ed.), Marshall’s 2017 Insects, their Natural History and Diversity (2nd ed.). Later, Huber and Poirier refined identifications and selected 420 specimens which they believed to represent distinct morphospecies to the Centre for Biodiversity Genomics at the University of Guelph for DNA barcoding (CO1-5P- 19 gene). These specimens were vouchered, photographed, and added to the Barcode of Life database (BOLD). BOLD can generate identifications by comparing query sequences from specimens with those in its reference library (Ratnasingham and Hébert 2007). Huber and Poirier used these to develop pictorial catalogs to help identify local species. Using the above-mentioned guides, pictorial catalogues, and online resources such as Ant Web, Bug Guide, and Facebook groups for fly and wasp identification, Honour’s student Alicja Muir further sorted specimens from four sampling periods (the weeks beginning on 17 July, 24 July, 6 August, and 28 August) to the lowest taxon she could (from species to order level). Two taxa, Megaselia and Platygastridae, were known for high diversity but difficult to sort, so an additional 285 specimens were submitted to BOLD. 2.4.4. Statistical analyses All statistical analyses were conducted in R version 4.2.2. I analyzed presence-absence data due to the biased nature of species abundances when using pitfall traps (Hohbein and Conway 2018). Most specimens were identified to genus or species. All genera and broader taxa were treated as unique identifications that I refer to as operational taxonomic units (OTUs). All OTU occurrences are recorded in (Table 5.1.2.) I assessed normality of data using histograms of residuals, plots of residuals vs fitted values, and normal Q-Q plots. For comparisons of categorical variables, I conducted ANOVA tests, except if the data’s residuals were non-normal, in which case, I used Kruskal-Wallis tests. 20 2.4.4.2. Baseline of species richness in Prince George Using the specaccum function in the R vegan package, I built species richness curves for OTU counts using the method “random”. I estimated the OTU richness for the regional species pool using the specpool function which calculates estimates using the Chao 2 index, first and second order jackknifing, and bootstrapping. 2.4.4.3. Comparison of land use types To ensure that sites within the same land use type were similar enough to be considered replicates, I used the geospatial data to build Bray-Curtis NMDS plots for each spatial scale. For each land use type, I compared the richness of species (OTUs), genera, and families of arthropods present. Species vary in their activity and capture rates in pitfall traps. Because I was comparing a broad range of taxa, with varying mobility, I avoided using abundance based diversity indices, as relative abundances from my trap yields would be unreliable (Topping and Sunderland 1992). For each taxonomic level, I estimated an upper value of richness where all OTUs were included, and a lower value of richness in which all broad taxa were excluded from the analysis (eg. if comparing genera, any categorization broader than genus would be excluded). I then conducted Kruskal-Wallis tests to compare richness values between land use types. 21 2.4.4.4. Community similarity I calculated the Bray-Curtis dissimilarities between sites, which when using presence absence data is equivalent to the Sørensen index. Unlike other indices, Bray-Curtis and Sorensen do not consider sites that don’t share a species (double zero problem) as similar to each other as sites that do share a species (Ramette 2007; Ricotta and Podani 2017). I drew ellipses around groups of sites belonging to the same land use type using the ordellipse function which calculates the standard error of the weighted centroids of for each group of sites. I conducted ANOSIM and PERMANOVA tests to see if the dissimilarity between land use types was significant at the OTU level. I also conducted ANOSIMs at the genus and family level where broader taxa were excluded. PERMANOVA is usually more sensitive to changes in community structure, and less affected by heterogenous dispersions between groups (Anderson and Walsh 2013). However, because they test different hypotheses, they can also be considered complimentary (Somerfield et al. 2021). These tests make no assumptions about the normality of the data and are compatible with Bray-Curtis dissimilarities (Anderson 2017). I conducted pairwise ANOSIM and PERMANOVA tests post-hoc to see which land use types were significantly different from each other. I adjusted p-values using the method fdr which calculates the false discovery rate for p-values (Jafari and Ansari-Pour 2019). 22 2.4.4.5. Habitat preference Urban areas typically favour open habitat species and habitat generalists. I categorized OTUs as preferring open, semi-open, closed habitats by reading about their habitat preferences in the literature (Table 5.1.3.). I classified OTUs that preferred a specific habitat, but likely occurred in others as “more open” or “more closed”. I classified OTUs as “semi-open” if they were edge species or preferred heterogenous habitats like prairie parkland. I would classify OTUs as “both” when they occurred in more than one habitat type without a clear preference. Because most OTUs had unknown habitat preferences, I could not perform a statistical analysis. I also tested whether any species were associated with different land use types using the rpb index from the indicspecies package. The function multipatt runs permutations of random communities to test the significance of species correlations with groups of sites (De Cáceres et al. 2010). 2.4.4.6. Geospatial modelling I modelled how upper estimates of species richness responds to seven geospatial land variables at all three spatial scales (50 m, 25 m, 10 m). I also included an additional variable called vegetation evenness, which measured how even or dominant different vegetation types (vegetable and flower garden, lawns, trees, and shrubs) were at each site using Pielou’s evenness index. Because the dataset was small, and there were many explanatory variables, including correlated ones, I used iterative forward selection where explanatory variables were only added to the final model if they were significant. I checked the normality of these models using Shapiro-Wilkes tests and histograms of residuals. While most Shapiro-Wilkes 23 tests had p-values larger than 0.05, suggesting normal distributions, visual inspection of the histograms showed some models had bimodal, right, or left-skewed distributions. Log and square-root transformations did not correct most of these distributions, so the variables were left untransformed. I also ran these models a second time using land use type as an interaction term (species richness = geospatial variable*land use type), as I expected arthropods from different land use types might respond differently to vegetative resources (e.g., grass could provide good habitat for industrial site-dwelling species but not necessarily for forest-dwelling species). I also fitted environmental variables to NMDS Bray-Curtis ordination plots to investigate community similarity using the function envfit from vegan at all three spatial scales. This function assesses the association between environmental variables and communities by modelling environmental variables as a function of ordination axis scores. It tests the significance of the relationship by randomly reordering the dependent variable through 10000 permutations and testing whether the original R2 is different from R2 distribution calculated through random permutations (Simpson 2018). 2.4.4.7. Functional diversity At its core, functional diversity involves grouping together ecologically similar species to better describe animal communities and how they respond to their environment (Wilson 1999). Functional groups refer to groups of species that perform the same roles in the ecosystem while guilds refer to groups of species that use similar resources, but both terms are often used interchangeably (Blondel 2003; 24 Cardoso et al. 2011). The simplest way to determine functional diversity is to measure the unique number of trait combinations present in a community. But researchers have developed a suite of more complex indices, many of which focus on trait richness (Mammola et al. 2021). Each index has strengths and weaknesses, so often it is important to include more than one index to ensure robust results (Petchey et al. 2009) or to measure complimentary facets of diversity (Legras et al. 2018). Functional traits can encompass various aspects of a species niche from their diet or habitat to their body size (Gu et al. 2016). I decided to focus on feeding guilds because these relate to some of the main ways arthropods contribute to the nutrient cycle and what resources they depend on to survive. I selected five guilds: herbivores, natural enemies (predators and parasites), decomposers, fungivores, and honeydew feeders. I chose very broad feeding guilds for three reasons. Oversplitting guilds could lead to some groups being overrepresented in the dataset or becoming redundant (e.g., many different feeding styles of herbivores or decomposers, if all were explicitly mentioned there would be far more herbivore and decomposer guilds than fungivore, and natural enemy guilds; or predators and parasitoids have very different lifestyles but essentially perform the same function of regulating other arthropod species). Including more traits would also likely make my functional diversity measures correlate too strongly with species richness and thus make it redundant (Petchey and Gaston 2006). And lastly, because many OTUs lacked taxonomic resolution, they were easier to sort into broader functional guilds 25 than into more specific ones. Narrower guilds would have led to a larger number of OTUs being excluded. Herbivores included arthropods that fed on any part of a plant and other nonanimal organism including pollen, nectar, lichen, mosses, and algae. Natural enemies included parasitoids and predators. Decomposers included arthropods that fed on any dead or rotting material including dead plants, carrion, or dung. Some decomposers may feed on decomposing microorganisms more so than the material under decomposition, but these were still considered decomposers due to their reliance on dead organic material. Many fungivores were also decomposers, so arthropods were only labelled as fungivores if fungi or fungal spores were explicitly mentioned to be a component of their diet. Honeydew feeders rely on plant juices obtained through a secondary species such as aphids, scale insects, or some fungi (Hardy 1988; Shaaban et al. 2021). Ants, in particular, are known for honeydew feeding, often even protecting the insects that supply these resources (Nielsen et al. 2010). Each OTU was ranked as having a strong (3), moderate (2), weak (1), or no (0) affinity for each guild, based on data from the literature (Table 5.1.1.). An OTU could belong to several different guilds, as many arthropods are omnivorous or feed on different food sources throughout their life cycle. Arthropods were ranked as three for a guild if it was their main feeding style, or main feeding style in their juvenile stage. Arthropods were ranked 2 if a food source was known to be important to their diet, or more important than other food sources, but not their main food source. They were ranked 1 for a food source if they fed on it facultatively, opportunistically, or 26 much less than other food resources. OTUs were ranked as 0 when the feeding style was extremely rare, not recorded in North America, or not present in that group. I excluded 18 OTUs of the 134 recorded due to uncertainty regarding their niches . Many adult insects, including flies and parasitoid wasps, feed on pollen, nectar, or honeydew. Often these were ranked as 1 for herbivory or honeydew feeders as feeding could be opportunistic, or the juvenile feeding stage was more strongly emphasized in the literature. Some OTUs had uncertain functional attributes. For broader taxa, I checked BOLD identifications from barcoded specimens, and referred to Canadian and provincial checklists (Campbell and Davies 1991; Campbell and Davies 1991; Hamilton 1998; Maw 2000; Beaulieu and Wheeler 2001; Lienhard 2018; Bennett et al. 2021; Langor and Langor 2022) to determine which species were most likely to occur in Prince George. OTUs that could represent two or more species with conflicting life histories were excluded from the functional analysis. When the feeding style was uncertain, OTUs were sorted into the guild of a closely related species, the most prevalent guild in a broader taxon, or the most likely guild based on ecological data. I used three different indices to measure functional diversity: the number of unique functional trait combinations (FRic), Petchey and Gaston’s 2002 FDPG , and the functional mean pairwise distances (FMPD) (Chapman et al. 2018). FDPG is a dendrogram based method where OTUs are sorted into a tree based on similarities in their functional traits. FDPG for a community is measured as total branch length required to include every member of the community (Petchey and Gaston 2002). FMPD measures the average distances between pairs of species from the same 27 community on a dendrogram (Tsirogiannis and Sandel 2014; Chapman et al. 2018). I used the FD package to calculate FRic, and methodology followed by Chapman et al. 2018 to calculate both FDPG and FMPD with the fundiv and picante R packages. To build my functional dendrogram, I used Gower’s distance which is appropriate for ordinal data (Podani and Schmera 2007; Mouchet et al. 2008). Several clustering algorithms can be used to build a tree, but the choice of algorithm can influence results. I selected algorithms recommended by Mouchet et al. 2008, used by Chapman et al. 2018 and Clark et al. 2012 in similar studies, and available in the fundiv package. I calculated the cophenetic correlations of all algorithms with my original distance matrices (Mouchet et al. 2008), and selected the unweighted pair group method with arithmetic mean (UPGMA) clustering algorithm as it scored the highest cophenetic correlation. I calculated the FDPG values for all sites and divided them by the FDPG value of a hypothetical community containing all OTUs to standardize them. I compared the functional diversity between land use types using a Kruskal Wallis test. I also tested their relationship with species richness using simple linear models. 2.4.4.8. Guild prevalence To analyze whether certain guilds were better represented in some land use types, I compared the proportion of each guild between land use types using Kruskal-Wallis tests. I used proportions to control for the effect of species richness between sites. For herbivores, decomposers, natural enemies, and fungivores, I only included OTUs that ranked moderate or higher in their feeding affinities to eliminate opportunistic and occasional feeders. I did not do this for honeydew feeders as most 28 OTUs in this guild were omnivorous and only ranked as weak. I also tested the association of different guilds with different land use types using richness data and the rpb index from the R package indicspecies. I expected guilds to have higher richness in land use types they were more strongly associated with (Panda et al. 2021). 2.4.4.9. Host specialization I ranked herbivores with a guild affinity of moderate or higher into three feeding categories: polyphagous OTUs that fed on many plant families (3), oligophagous OTUs that fed within a single family (2), and monophagous OTUs that feed on a single genus (1) (Table 5.1.4.). I included 33 taxa whose host specialization was known out of 39 taxa that were known to be herbivorous (Table 3.4.). I performed Kruskal Wallis tests to compare the proportions of each host specialization category between land use types and corrected the p-values using the false detection rate function fdr. I also calculated the mean host specialization of each site and compared it between land use types. 2.5. RESULTS 2.5.1. Land use type classification These figures plot sites in NMDS space based off their geospatial dissimilarities (differences in area of grass, hard surfaces, trees, shrubs, buildings, and gardens) at the 10 m, 25 m, and 50 m scales. These results show that sites within the same land use type are similar in their geospatial composition but distinct from other land use types at every spatial scale. 29 Figure 2.2. NMDS Bray-Curtis dissimilarity plots where points represent sites and ellipses represent the standard error of weighted centroids. Plots were constructed using vegan in R and the function metaMDS, from geospatial data at the 10 m, 25 m, and 50 m scales. The data were standardized with square-root transformations, then the Wisconsin double standardization. Plot A represents sites at the 10 m scale (dissimilarity = “bray”, k = 3, permutations = 20, stress = 0.0031). Plot B represents sites at the 25 m scale (distance = Bray, k = 3, permutations = 20, stress = 0.014). Plot C represents sites at the 50 m scale (distance = bray, k = 3, permutations = 20, stress = 0.0027). Note: Some points overlap in figures. 2.5.2. Species baseline 2.5.2.1. Data quality In 2015, 8645 specimens were sorted into 134 operational taxonomic units (OTUs), where OTUs represented the lowest taxonomic level to which a species could be identified. Thirty-six percent of OTUs were identified to species (5% of specimens), 35% to genus (86% of specimens), and 25% to family (6% of specimens). The proportion of OTUs identified to genus or lower was weakly correlated with OTU richness (Pearson’s, t = 2.3, df = 10, cor = 0.59, p=0.044), but 30 these proportions did not differ significantly between land use types (ANOVA, df = 2,9, F = 3.45, p = 0.077). While taxonomic resolution may have affected species richness values, the resolution was similar across greenbelt, industrial, and residential sites. Figure 2.3. The OTU richness at each site. Different colors indicate the proportion of OTUs at each site that represent species-, genus-, family-, and order-level identifications. 2.5.2.2. Richness estimates Overall, 134 morphological OTUs were detected across 89 genera and 63 families from initial barcoding and sorting with guides and pictorial catalogs. Barcoded specimens represented 118 molecular OTUs, with 33 OTUs belonging to Megaselia. BOLD records included an additional 130 OTUs that were sampled outside the initial collection periods Alicja Muir studied (17 July, 24 July, 6 August, 28 August) for a total of 248 molecular OTUs recorded in Prince George. Given the position of the curve in Figure 2.4., this likely represents most of the OTUs that can 31 be found through morphological sorting alone. Morphological OTU estimates for the regional species pool ranged from 161 +/- 10 (bootstrap index) to 246 +/- 38 (chao 2 index) indicating that between 54% to 83% OTUs in the regional species pool were sampled. Figure 2.4. The OTU accumulation curve of OTUs detected from the 8645 specimens across 12 sites and four sampling periods. Using the specaccum (method=random) package from vegan in R, sites were added randomly in 100 permutations of the data to find the mean species accumulation curve. Dots represent the mean number of OTUs found per number of sites sampled (+/- the standard deviation in blue). A total 134 OTUs were sampled across the sites and collecting periods. Table 2.1. Estimates of OTU richness for the regional species pool and each land use type, along with their standard error for the chao 2 index, jackknife 1, jackknife2, and bootstrap estimators. Land use type All greenbelt industrial residential OTU richness 134 57 83 68 Chao 2 Jack 1 Jack 2 Bootstrap 246 +/- 38 91 +/- 16 141 +/- 24 100 +/- 15 198 +/- 23 82 +/- 16 119 +/- 22 95 +/- 20 241 94 138 108 161 +/- 11 68 +/- 8 99 +/- 10 80 +/- 10 32 2.5.3. Richness between land use types When including all OTUs at the finest taxonomic level, industrial sites had the highest overall mean richness, a residential site had the highest absolute richness value, and richness did not differ significantly between land use types (KruskalWallis, ߯ ଶ = 4.29, df = 2, p = 0.12). Upper and lower estimates for genus and family richness did not significantly differ either. Table 2.2. Comparisons of richness values between land use types where upper estimates of richness include all taxa and lower estimates exclude broader taxa. (df = degrees of freedom, res. df. = residual degrees of freedom, p = p-value). Response variable Species richness upper estimate Species richness lower estimate Genus richness upper estimate Genus richness lower estimate Family richness upper estimate Family richness lower estimate Type of test Kruskal Wallis Kruskal Wallis Kruskal Wallis ANOVA Kruskal Wallis ANOVA F-statistic ߯ଶ res. df. 4.29 1.53 4.29 3.47 9 4.02 1.93 9 df. 2 2 2 2 2 2 p 0.12 0.47 0.12 0.08 0.13 0.20 33 Figure 2.5. Dot plots representing upper (left column), and lower (right column) estimates for species, genus, and family richness between land use types. Upper estimates include all taxa. Lower estimates excluded broad taxa. 34 2.5.4. Geospatial land feature modelling Contrary to what was expected, neither the total area of vegetation, nor vegetation evenness predicted species richness. Of the 54 models tested, only three proved significant, the area covered by trees at the 25 m scale, and the area of trees and grass at the 10 m scale. The first model was non-normal. The last two were strongly correlated (Pearson’s correlation = -0.81, t = -4.3, df = 10, p = 0.0015), and so were not combined into a single, final model. Figure 2.6. Scatterplots of species richness as a function of total vegetated area (A) and as a function of Pielou’s evenness (B) for different vegetation types within a 10 m radius. 35 Table 2.3. The results of simple linear regressions where the total vegetated area (tot. veg.) and Pielou’s evenness vegetation type (veg. evenness) are used to predict species richness (SR). (Adj. R2 = Adjusted R2) Equation SR~ total veg. SR ~ total veg. SR ~ total veg. SR ~ veg. evenness SR ~ veg. evenness SR ~ veg. evenness Scale (m) Slope 50 25 10 50 25 10 -0.0017 -0.0074 -0.03 -4.87 -6.57 3.01 FRes. statistic df 2.88 10 3.47 10 0.74 10 0.05 10 0.52 10 0.14 10 df Adj. R2 p 1 1 1 1 1 1 0.15 0.18 -0.024 -0.095 -0.046 -0.085 0.12 0.092 0.41 0.83 0.49 0.72 Table 2.4. The results of two simple linear regressions; species richness (SR) as a function of grass area, and species richness as a function of forest area, both at a ten-meter scale. Both models are significant, but grass area negatively correlates with tree area, so these results are not independent. Equation SR ~ lawn area SR ~ tree area Scale 10 m 10 m Slope 0.079 -0.04 F-statistic 10.31 5.41 Res. df. 10 10 df 1 1 Adj. R2 0.46 0.29 p 0.0093 0.042 Figure 2.7. Linear regression plots of upper OTU richness estimates as a function of grass area (A) and forest area (B) within ten meters of a site. 36 2.5.5. Community similarity Greenbelt and industrial arthropod communities were expected to be distinct from each other because they represented two extremes of vegetation types (tall vegetation, shaded, lots of woody debris vs. short, dense, open vegetation). Overall, industrial, residential, and greenbelt communities were significantly dissimilar from each other at the OTU, genus, and family level (Table 2.5.). The area of trees, grass, hard-surfaces, and buildings were all significantly associated with community composition (Figure 2.9.). When OTUs were pooled for each land use type, industrial and residential communities shared more OTUs with each other than with greenbelt sites. However, industrial communities had the greatest number of OTUs, and the most unique OTUs overall (Figure 2.8.). Geographic distance between sites did not correlate with community dissimilarity (Pearson’s r = 0.084, t = 0.68, df = 64, p = 0.5). 37 Figure 2.8. A Venn Diagram of the number of OTUs that are unique or shared between land use types (constructed from pooled data for each land use type). The percentages indicate what proportion of the total OTU pool each category makes up. 38 Figure 2.9. Bray Curtis ordination (dissimilarity = "bray", dimensions = 4, permutations = 10000, stress = 0.052) plots arthropod communities in NMDS space where points represent sites, and color indicates land use type (meta.mds function from vegan in R). Ellipses were drawn around sites of the same land use type using the standard error of weighted centroids. Residential sites overlap more strongly with industrial sites, indicating they have more similar species communities than greenbelt sites which are more distinct from the other two. Vectors indicate the relationship of environmental variables to communities using the envfit function from the vegan package. Longer arrows indicate a stronger relationship. These were tested at the 10 m (A), 25 m (B), and 50 m (C) scale with 10000 permutations. Tree area and grass area correlate with each other at the 10 m scale, so these results are not independent. 39 Table 2.5. Results of ANOSIM (ANO) and PERMANOVA (PERM) tests which compared the similarity of greenbelt, residential, and industrial communities. All analyses were conducted using Bray-Curtis dissimilarity with 10000 permutations. (Gb. = Greenbelt, Ind. = Industrial, Res. = Residential) Test ANO Res. df R df. All OTU ANO 0.62 0.0002 Gb & Ind. OTU ANO 0.74 0.03 0.03 Res. & Gb. OTU ANO 0.65 0.03 0.03 Res. & Ind. OTU ANO 0.57 0.03 0.03 All OTU PERM Gb & Ind. OTU PERM 0.03 Res. & Gb. OTU PERM 0.03 Res. & Ind. OTU PERM 0.03 All Genus ANO 0.68 0.0002 All Family ANO 0.51 0.0003 9 2 2 R 0.32 Res. R2 0.68 FStatistic fdr adj. p Taxon Communities Level 2.16 p 0.0002 2.5.6. Habitat preference 40 Based on abundance data, Only Megaselia and Lasius were significantly associated with greenbelt and residential habitats respectively (rpb = 0.58, p = 0.03; rpb = 0.97, p = 0.01). However, species within these genera can have differing habitat preferences. A few other OTUs were strongly associated with different land use types, but a larger sample size would be needed to confirm these preferences (Table 3.3.). No analyses could be performed for habitat preference as the majority of OTUs’ niches remain unknown. However most known OTUs seem to prefer semiopen habitat or are habitat generalists found in more than one type of environment. Closed habitat taxa are mostly found at greenbelt sites and open habitat taxa are mostly found at residential and industrial sites. Figure 2.10. Bar plot of the habitat preference for different OTUs, where bars represent OTU richness and different colors represent habitat preference as described in the literature. 41 2.5.7. Functional diversity Although urbanization was expected to negatively impact functional diversity, diversity was similar across land use types, regardless of the index used. Species richness significantly predicted FDpg and FRic but not FMPD, suggesting results are robust. Table 2.6. The results of three Kruskal Wallis tests that compared functional richness values of three different indices against land use type. Statistical Test Kruskal-Wallis Kruskal-Wallis Kruskal-Wallis Index FD(pg) F(Ric) FMPD ߯ଶ 3.23 1.61 1.42 df. 2 2 2 p 0.20 0.45 0.49 Table 2.7. The results of three simple linear regressions that model functional richness as a function of species richness. FDpg and FRic are both strongly affected by species richness while FMPD is not. 42 Index FDPG FRic FMPD Slope 0.0089 0.242 0.0021 F-statistic 8.36 37.38 0.65 Res. df. 10 10 10 df. 1 1 1 Adj. R2 0.40 0.77 -0.03 p 0.016 0.00011 0.44 43 p < 0.001, Adj R2 = 0.77 p = 0.016, Adj R2 = 0.40 Adj R 0.40 2 p = 0.44, Adj R2 = -0.03 Adj R2 0.40 Figure 2.11. The three box and whiskers plots above compare the median functional richness values (FRic, FDpg, FMPD) of different land use types, while the regression plots below show functional richness as a function of OTU richness. 44 2.5.8. Guild prevalence Urbanization was expected to select for lower trophic positions. Residential and industrial areas did have higher proportions of herbivores and lower proportions of natural enemies, and these differences were significant for herbivores in residential areas compared to greenbelt areas (߯ ଶ = 6.99, df = 2, p = 0.03). The proportions of decomposers, fungivores, and honeydew feeders were statistically similar across land use types. Herbivore richness was most strongly associated with industrial and residential sites (rpb = 0.87, adjusted p = 0.026), while no other guilds correlated significantly with land use types (Table 2.9.). Figure 2.12. The box and whiskers plots compare the proportion taxa bellowing to herbivores (top left), natural enemies (top right), and decomposers (bottom left) out of all potential taxa at each site between land use types. 45 Table 2.8. The results of Kruskal-Wallis and Dunn tests comparing the proportions of different guilds against land use types. (Herb. = Herbivore, Nat. En. = Natural Enemy, Decom = Decomposer, fdr adj. p =fdr adjusted p-value) Statistical Test Kruskal-Wallis Dunn-Test Dunn-Test Dunn-Test Kruskal-Wallis Kruskal-Wallis Guild Herb. Herb. Herb. Herb. Nat. En. Decom. Comparison Gb. – Ind. Gb. – Res. Ind. – Res. ߯ଶ 6.97 6.97 6.97 6.97 5.10 3.27 df 2 2 2 p 0.03 0.03 0.005 0.3 0.08 0.2 fdr adj. p 0.05 0.02* 0.2 Table 2.9. The associations of guilds with land use types. Correlations were calculated using the rpb index from the indicspecies package in R (permutations = 10000). Zeroes and ones indicate positive or no association between guilds and land use types. rpb is the correlation between guilds and groups of sites for their strongest association (single or multiple land use types). P-values were adjusted using the fdr method. Guild Herbivores Natural Enemies Decomposers Fungivores Honeydew Feeders Gb. 0 0 0 0 0 Ind. 1 1 1 1 1 Res. 1 0 1 0 0 rpb 0.87 0.33 0.62 0.34 0.35 p 0.005 0.58 0.10 0.68 0.65 fdr. adj. p 0.026* 0.68 0.24 0.68 0.68 46 2.5.9. Host specialization Industrial areas were expected to have higher proportions of polyphagous OTUs. Overall, there were higher proportions of specialists at greenbelt sites, and more generalists at residential and industrial sites, but there were no significant differences. Generalists were the most common host specialization level, followed by oligophagous taxa, suggesting that a broad diet may be beneficial overall in urban environments. Mean host specialization also showed no significant differences between land use types. A larger sample size may be needed to draw more conclusive results. Figure 2.13. The proportions of different host specialization categories in different colors over the total number of herbivores collected for each site. 47 Figure 2.14. Box and whiskers plot of mean host specialization by land use type, calculated using the CWM.type method, in dbFD function from the FD package in R. Specialists were ranked as 1, oligophages as 2, and generalists ranked as 3. Table 2.10. The results of Kruskal-Wallis tests for the proportions of generalist, oligophagous, and specialist OTUs compared between land use types, as well as the means between land use types. Statistical Test KruskalWallis KruskalWallis KruskalWallis KruskalWallis Dependent Variable Proportion of Generalists Proportion of Oligophages Proportion of Specialists Mean host specialization ߯ଶ df p fdr adj. p 1.44 2 0.49 0.49 6.39 2 0.04 0.12 4.35 2 0.11 0.17 2.23 2 0.33 48 2.6. DISCUSSION While many studies highlight the negative impacts of urbanization on arthropod communities (Vergnes et al. 2014; Chatelain et al. 2023), these findings suggest that land use type had no strong effects on diversity. However, communities from different land use types were distinct, highlighting how many different styles of land management can foster high gamma diversity across a city (Sattler et al. 2011). McIntyre et al. 2001 found similar results where land use type did not affect taxon richness but did alter community composition. Sattler et al. 2010 also found little difference in arthropod richness along urban gradients, which they attributed to the fine-scale habitat heterogeneity that cities create. Urban habitats filter species from the regional species pool based on their ecological traits (Aronson et al. 2016) and are typically considered harsh environments in which only few species can survive (Sol et al. 2014). But urban conditions can also promote higher than average diversity, especially when food availability is high. Often, well-adapted, non-native species are introduced (Gippet et al. 2019), or early seral species colonize urban habitats (McKinney 2006). While urbanization may be cause for concern on a large scale, the conservation value of industrial and residential areas should not be ignored as they can support taxonomically diverse arthropod communities. Although increased total vegetation and vegetation evenness would be expected to positively affect biodiversity, neither of these variables had any effect. Lawn at the ten-meter scale was found to positively affect arthropod richness, while tree area decreased it. This could be due in part to greenbelt sites tending to have 49 lower species richness than grassier industrial and residential sites. However, of the taxa with known habitat preferences, only a handful preferred closed habitats. Many greenbelt species appeared to be habitat generalists or preferred semi-open habitats. In other land use types, taxa tended to be habitat generalists or open habitat species. Overall, most OTUs were positively associated with industrial (53%) and residential (34%) areas. Urban areas often favour habitat generalists and open-habitat species (Magura et al. 2010; Olivier et al. 2015; Magura et al. 2020). Since lawn is one of the most common vegetative characteristics in cities (Proske et al. 2022), and most taxa in Prince George prefer or tolerate lawns, it follows that increasing lawn area, and consequently, decreasing tree area, could promote higher OTU richness, potentially even in forested areas. Urban forests are often invaded by open-habitat species from the surrounding landscape, especially as urban forests tend to be drier and warmer than natural forests (Magura et al. 2020). Lawn at the ten-meter scale might also act as a better proxy for habitat patch size in residential and industrial sites than at larger spatial scales as it is less likely to be measuring vegetation separated by fences, buildings, roads, or other physical obstacles. Thus, increases in lawn area may relate to increases in overall habitat size. Note that lawns in this study tended to be lightly managed with infrequent mowing and little to no use of herbicides and pesticides. Functional diversity was also similar between land use types, regardless of the index used. Because abundance was not considered, only functional richness could truly be evaluated, but this suggests that taxa fill the same basic niches in 50 various land use types, and members of those communities are, on average, equally dissimilar from each other. Both FRic and FDPG were significantly related to species richness, as communities with larger species pools are more likely to have functionally different species (Petchey et al. 2009). However, a strong relationship with species richness can make it hard to disentangle the effect of richness from the effect of functional differences within a community (Dalerum et al. 2012). FMPD was not significantly affected by species richness but was still similar across land use types, suggesting these results are robust. While arthropod communities in different land use types may offer the same range of ecosystem services, different guilds may be better represented in different land use types. Proportions of honeydew feeders, fungivores, and decomposers were similar between land use types. But greenbelt communities supported higher proportions of natural enemies while residential areas harboured significantly higher proportions of herbivores. Herbivores were also significantly associated with industrial and residential sites. Natural enemies are especially vulnerable to environmental disturbance (Cardoso et al. 2011). They often require a wide array of resources to complete their life cycles, from nectar to prey sources (Corcos et al. 2019). Natural enemies need larger, more well-connected habitats than their prey (Kruess and Tscharntke 1994), and urban features like roads and walls can represent dispersal barriers (Peralta et al. 2011). Greenbelt sites represented the largest, most continuous habitat patches as they had the highest total vegetation, and smallest areas of buildings and hard surfaces. 51 Greenbelt sites were also mainly unmanaged, which could better support longer food chains. Rigal et al. 2018 found intensive pasture selected for piercecutting herbivores compared to more diverse assemblages in natural forests. Likewise Hironaka and Koike 2013 found semi-natural grasslands better supported large carnivorous arthropods compared to managed grasslands. Additionally, decreasing native plant richness and habitat complexity at urbanized sites can reduce predator pressure (Raupp et al. 2010), which may explain higher proportions of herbivores at industrial and residential sites. Yet even if some groups are underrepresented at various land use types, the mosaic nature of cities helps compensate for this as different functional groups can be promoted in different land use types (Pinto et al. 2021). Urbanization tends to select for generalist species (Clavel et al. 2011), but host specialization was similar for herbivores across land use types. This suggests herbivores face similar selection pressures for this trait regardless of land management style, however, a larger sample size would be needed to confirm these results. On a global scale, urbanization threatens biodiversity by removing an increasing number of vulnerable species from regional pools (Clavel et al. 2011), and homogenizing remaining communities (McKinney 2006). As cities sprawl, novel habitats become increasingly important for protecting biodiversity as natural areas alone prove insufficient (Lundholm and Richardson 2010). On a local level, urban planners and landowners can maximize biodiversity by promoting conservation in land use types that are typically overlooked. If industrial and residential areas are as 52 taxonomically and functionally diverse as urban forests, then they should be managed to preserve their ecological value. Further research could explore ways to promote arthropod diversity in urban land use types that I did not explore in this study, such as reduced mowing, intentionally increasing plant diversity in developed areas, maintaining current greenbelt sites, and planning for greenbelt in new residential and industrial areas (Proske et al. 2022). Future studies could also expand sampling for pollinators or flying/overstory arthropods which also contribute to urban ecosystem services and consider the effect of geospatial variables at larger scales (eg. 500 meters) to see how arthropods respond to landscape-level urbanization. 2.7. REFERENCES Anderson MJ. 2017. Permutational Multivariate Analysis of Variance (PERMANOVA). In: Wiley StatsRef: Statistics Reference Online. p. 1–15. [accessed 2023 Aug 31]. https://doi.org/10.1002/9781118445112.stat07841. Anderson MJ, Walsh DCI. 2013. PERMANOVA, ANOSIM, and the Mantel test in the face of heterogeneous dispersions: What null hypothesis are you testing? Ecol. Monogr. 83(4):557– 574. Aronson MFJ, Nilon CH, Lepczyk CA, Parker TS, Warren PS, Cilliers SS, Goddard MA, Hahs AK, Herzog C, Katti M, et al. 2016. Hierarchical filters determine community assembly of urban species pools. Ecology (Durham). 97(11):2952–2963. doi:10.1002/ecy.1535. Beaulieu F, Wheeler TA. 2001. Inventaire des espèces de brachycères (Diptera) des prés de laîches (Cyperaceae, Carex) de la Réserve nationale de faune du lac Saint-François, Québec. Fabreries. 26(2):57–74. Beninde J, Veith M, Hochkirch A. 2015. Biodiversity in cities needs space: A meta-analysis of factors determining intra-urban biodiversity variation. Ecol. Lett. 18(6):581–592. doi:10.1111/ele.12427. Bennett R, Copley D, Copley C. 2021. Checklist of the spiders (Araneae) of British Columbia. J. Entomol. Soc. B.C. 118:25–52. Blondel J. 2003. Guilds or functional groups: Does it matter? Oikos. 100:223–231. doi:10.1034/j.1600-0706.2003.12152.x. 53 Campbell J, Davies A. 1991. Family Staphylinidae rove beetles. In: Checklist of beetles of Canada and Alaska. Research Branch Agriculture Canada Publication. Ottawa, Ontario. p. 86–124. Cardoso P, Pekár S, Jocqué R, Coddington JA. 2011. Global patterns of guild composition and functional diversity of spiders. PLOS One. 6(6):e21710. doi:10.1371/journal.pone.0021710. Chapman PM, Tobias JA, Edwards DP, Davies RG. 2018. Contrasting impacts of land-use change on phylogenetic and functional diversity of tropical forest birds. J. Appl. Ecol. 55(4):1604–1614. doi:10.1111/1365-2664.13073. Chatelain M, Rüdisser J, Traugott M. 2023. Urban-driven decrease in arthropod richness and diversity associated with group-specific changes in arthropod abundance. Front. Ecol. Evol. 11. https://www.frontiersin.org/articles/10.3389/fevo.2023.980387. City of Prince George. 2023. Our Infrastructure True Story. City of Prince George. https://www.princegeorge.ca/city-hall/infrastructure/our-infrastructure-story. Clark CM, Flynn DFB, Butterfield BJ, Reich PB, Cahill JFE. 2012. Testing the link between functional diversity and ecosystem functioning in a Minnesota grassland experiment. PLOS One. 7(12). doi:10.1371/journal.pone.0052821. Clavel J, Julliard R, Devictor V. 2011. Worldwide decline of specialist species: toward a global functional homogenization? Front. Ecol. Environ. 9(4):222–228. Coady C, Picketts I. 2012. Implementing Climate Change Adaptation in Prince George, BC, Volume 3. Prince George, British Columbia Implementing Climate Change Adaptation in Prince George, BC Report No.: 3. [accessed 2023 Aug 29]. https://www.princegeorge.ca/media/142. Concepción ED, Moretti M, Altermatt F, Nobis MP, Obrist MK. 2015. Impacts of urbanisation on biodiversity: the role of species mobility, degree of specialisation and spatial scale. Oikos. 124(12):1571–1582. doi:10.1111/oik.02166. Corcos D, Cerretti P, Caruso V, Mei M, Falco M, Marini L. 2019. Impact of urbanization on predator and parasitoid insects at multiple spatial scales. Chapman M G., editor. PLOS One. 14(4):e0214068–e0214068. doi:10.1371/journal.pone.0214068. Cramer MJ, Willig MR. 2005. Habitat heterogeneity, species diversity and null models. Oikos. 108(2):209–218. Dalerum F, Cameron EZ, Kunkel K, Somers MJ. 2012. Interactive effects of species richness and species traits on functional diversity and redundancy. Theor. Ecol. 5(1):129–139. doi:10.1007/s12080-010-0104-y. De Cáceres M, Legendre P, Moretti M. 2010. Improving indicator species analysis by combining groups of sites. Oikos 119: 1674-1684. doi:10.1111/j.1600-0706.2010.18334.x. Gippet JM, Liebhold AM, Fenn-Moltu G, Bertelsmeier C. 2019. Human-mediated dispersal in insects. Curr. Opin. Insect Sci. 35:96–102. doi:10.1016/j.cois.2019.07.005. 54 Gu J, Zhou J, Wilson M, Jia K, Lv K, Xu Z. 2016. Species diversity and functional diversity of insects in Wuxijiang National Wetland Park, East China. Acta Ecol. Sin. 36(5):386–391. doi:10.1016/j.chnaes.2016.07.002. Hamilton KGA. 1998. The species of the North American leafhoppers Ceratagallia Kirkaldy and Aceratagallia Kirkaldy (Rhynchota: Homoptera: Cicadellidae). Can. Entomol. 130(4):427–490. doi:10.4039/Ent130427-4. Hardy TN. 1988. Gathering of fungal honeydew by Polistes spp. (Hymenoptera: Vespidae) and potential transmission of the causal ergot fungus. Fla. Entomol. 71(3):374–376. doi:10.2307/3495447. Hironaka Y, Koike F. 2013. Guild structure in the food web of grassland arthropod communities along an urban-rural landscape gradient. Écoscience. 20(2):148–160. Hohbein RR, Conway CJ. 2018. Pitfall traps. Wildl. Soc. Bull. 42(4):597–606. Jafari M, Ansari-Pour N. 2019. Why, when and how to adjust your p values? Cell J. 20(4):604–607. doi:10.22074/cellj.2019.5992. Kruess A, Tscharntke T. 1994. Habitat fragmentation, species loss, and biological control. Science. 264(5165):1581–1584. doi:10.1126/science.264.5165.1581. Langor DW, Langor SD. 2022. The biota of Canada: checklist of the centipedes of Canada (Myriapoda: Chilopoda). Can. Entomol. 154(1):e8. doi:10.4039/tce.2021.58. Legras G, Loiseau N, Gaertner J-C. 2018. Functional richness: Overview of indices and underlying concepts. Acta Oecol. 87:34–44. doi:10.1016/j.actao.2018.02.007. Lienhard C. 2018. Country Checklists of the Psocoptera species of the World Extracted from Lienhard & Smithers, 2002: “Psocoptera (Insecta)-World Catalogue and Bibliography.” Sapporo, Japan: The Psocidologists’ Newsletter. https://www.researchgate.net/publication/329281524_COUNTRY_CHECKLISTS_OF_THE_ PSOCOPTERA_SPECIES_OF_THE_WORLD_EXTRACTED_FROM_LIENHARD_SMITHE RS_2002_PSOCOPTERA_INSECTA_-_WORLD_CATALOGUE_AND_BIBLIOGRAPHY. Lundholm JT, Richardson PJ. 2010. Habitat analogues for reconciliation ecology in urban and industrial environments. J. Appl. Ecol. 47(5):966–975. Magura T, Ferrante M, Lövei GL. 2020. Only habitat specialists become smaller with advancing urbanization. Glob. Ecol. Biogeogr. 29(11):1978–1987. doi:10.1111/geb.13168. Magura T, Horváth R, Tóthmérész B. 2010. Effects of urbanization on ground-dwelling spiders in forest patches, in Hungary. Landsc. Ecol. 25(4):621–629. doi:10.1007/s10980009-9445-6. Mammola S, Carmona CP, Guillerme T, Cardoso P. 2021. Concepts and applications in functional diversity. Funct. Ecol. 35(9):1869–1885. doi:10.1111/1365-2435.13882. Maw HEL. 2000. Checklist of the Hemiptera of Canada and Alaska. NRC Research Press. https://cdnsciencepub.com/doi/book/10.1139/9780660181653. 55 McIntyre NE, Rango J, Fagan WF, Faeth SH. 2001. Ground arthropod community structure in a heterogeneous urban environment. Landsc. Urban Plann. 52(4):257–274. doi:10.1016/S0169-2046(00)00122-5. McKinney ML. 2006. Urbanization as a major cause of biotic homogenization. Biol. Conserv. 127(3):247–260. doi:10.1016/j.biocon.2005.09.005. Mouchet M, Guilhaumon F, Villéger S, Mason NWH, Tomasini J-A, Mouillot D. 2008. Towards a consensus for calculating dendrogram-based functional diversity indices. Oikos. 117(5):794–800. doi:10.1111/j.0030-1299.2008.16594.x. Nielsen C, Agrawal AA, Hajek AE. 2010. Ants defend aphids against lethal disease. Biology letters (2005). 6(2):205–208. doi:10.1098/rsbl.2009.0743. Olivier T, Schmucki R, Fontaine B, Villemey A, Archaux F. 2015. Butterfly assemblages in residential gardens are driven by species’ habitat preference and mobility. Landsc. Ecol. 31(4):865–876. doi:10.1007/s10980-015-0299-9. Panda BP, Prusty BAK, Panda B, Pradhan A, Parida SP. 2021. Habitat heterogeneity influences avian feeding guild composition in urban landscapes: evidence from Bhubaneswar, India. Ecol. Process. 10(1):31. doi:10.1186/s13717-021-00304-6. Peralta G, Fenoglio MS, Salva A. 2011. Physical barriers and corridors in urban habitats affect colonisation and parasitism rates of a specialist leaf miner. Ecol. Entomol. 36(6):673– 679. doi:10.1111/j.1365-2311.2011.01316.x. Petchey O, O’Gorman E, Flynn D. 2009. A functional guide to functional diversity measures. In: Biodivers. Ecosyst. Funct. Hum. Wellbeing Ecol. Econ. Perspect. Vol. 85. Petchey OL, Gaston KJ. 2002. Functional diversity (FD), species richness and community composition. Ecol. Lett. 5(3):402–411. doi:10.1046/j.1461-0248.2002.00339.x. Petchey OL, Gaston KJ. 2006. Functional diversity: back to basics and looking forward. Ecol Lett. 9(6):741–758. doi:10.1111/j.1461-0248.2006.00924.x. Pinto CM, Pairo PE, Bellocq MI, Filloy J. 2021. Different land-use types equally impoverish but differentially preserve grassland species and functional traits of spider assemblages. Sci. Rep. 11(1):10316. doi:10.1038/s41598-021-89658-7. Podani J, Schmera D. 2007. How should a dendrogram-based measure of functional diversity function? A rejoinder to Petchey and Gaston. Oikos. 116(8):1427–1430. doi:10.1111/j.0030-1299.2007.16160.x. Prince George Area Industrial Land Profile. 2008. PGArea_Industrial_Lands_Profile.pdf. Proske A, Lokatis S, Rolff J. 2022. Impact of mowing frequency on arthropod abundance and diversity in urban habitats: A meta-analysis. Urban For. Urban Green. 76:127714. doi:10.1016/j.ufug.2022.127714. Ramette A. 2007. Multivariate analyses in microbial ecology. FEMS Microbiol. Ecol. 62(2):142–160. doi:10.1111/j.1574-6941.2007.00375.x. 56 Ratnasingham S, Hébert PDN. 2007. BOLD: The Barcode of Life Data System ( http://www.barcodinglife.org). Mol. Ecol. Notes. 7(3):355–364. doi:10.1111/j.14718286.2007.01678.x. Raupp MJ, Shrewsbury PM, Herms DA. 2010. Ecology of herbivorous arthropods in urban landscapes. Annu Rev Entomol. 55(1):19–38. doi:10.1146/annurev-ento-112408-085351. Ricotta C, Podani J. 2017. On some properties of the Bray-Curtis dissimilarity and their ecological meaning. Ecol. Complex. 31:201–205. doi:10.1016/j.ecocom.2017.07.003. Rigal F, Cardoso P, Lobo JM, Triantis KA, Whittaker RJ, Amorim IR, Borges PAV. 2018. Functional traits of indigenous and exotic ground-dwelling arthropods show contrasting responses to land-use change in an oceanic island, Terceira, Azores. Divers. Distrib. 24(1/2):36–47. Sattler T, Duelli P, Obrist MK, Arlettaz R, Moretti M. 2010. Response of arthropod species richness and functional groups to urban habitat structure and management. Lands. Ecol. 25(6):941–954. doi:10.1007/s10980-010-9473-2. Sattler T, Obrist MK, Duelli P, Moretti M. 2011. Urban arthropod communities: Added value or just a blend of surrounding biodiversity? Landsc. Urban Plann. 103(3):347–361. doi:10.1016/j.landurbplan.2011.08.008. Shaaban B, Seeburger V, Schroeder A, Lohaus G. 2021. Suitability of sugar, amino acid, and inorganic ion compositions to distinguish fir and spruce honey. Eur. Food Res. Technol. 247(4):879–888. doi:10.1007/s00217-020-03671-0. Simpson G. 2018. vector fit interpretation NMDS. Cross Validated. [accessed 2023 Dec 14]. https://stats.stackexchange.com/questions/56427/vector-fit-interpretation-nmds. Sol D, González-Lagos C, Moreira D, Maspons J, Lapiedra O. 2014. Urbanisation tolerance and the loss of avian diversity. Ecol. Lett. 17(8):942–950. doi:10.1111/ele.12297. Somerfield PJ, Clarke KR, Gorley RN. 2021. Analysis of similarities (ANOSIM) for 2-way layouts using a generalised ANOSIM statistic, with comparative notes on Permutational Multivariate Analysis of Variance (PERMANOVA). Austral Ecol. 46(6):911–926. doi:10.1111/aec.13059. Statistics Canada. 2023. Census Profile. 2021 Census of Population. [accessed 2023 Aug 27]. https://www12.statcan.gc.ca/census-recensement/2021/dp-pd/prof/index.cfm?Lang=E. Timberline Forest Inventory Consultants. 2006. Community Forest of Prince George: Management Plan. [accessed 2023 Aug 29]. https://treecanada.ca/wpcontent/uploads/2017/11/City-of-prince-George.pdf. Topping CJ, Sunderland KD. 1992. Limitations to the use of pitfall traps in ecological studies exemplified by a study of spiders in a field of winter wheat. J. Appl. Ecol. 29(2):485–491. doi:10.2307/2404516. Tóth Z, Hornung E, Szlavecz K. 2021. Urban effects on saprophagous macroarthropods are mainly driven by climate: A global meta-analysis. Sci. Total Environ. 797:149182. doi:10.1016/j.scitotenv.2021.149182. 57 Tsirogiannis C, Sandel B. 2014. Computing the skewness of the phylogenetic mean pairwise distance in linear time. Algorithms for Molecular Biology. 9(1):15. doi:10.1186/1748-7188-915. Turrini T, Knop E. 2015. A landscape ecology approach identifies important drivers of urban biodiversity. Glob. Change Biol. 21(4):1652–1667. doi:10.1111/gcb.12825. Vergnes A, Pellissier V, Lemperiere G, Rollard C, Clergeau P. 2014. Urban densification causes the decline of ground-dwelling arthropods. Biodivers. Conserv. 23(8):1859–1877. doi:10.1007/s10531-014-0689-3. Wilson JB. 1999. Guilds, functional types and ecological groups. Oikos. 86(3):507–522. doi:10.2307/3546655. 58 Chapter three: The abundance and diversity of scuttle flies in Prince George 3.1. INTRODUCTION While concerns mount over declining insect abundance and diversity (Forister et al. 2019), few groups remain as species-rich, but as simultaneously understudied as the scuttle fly family (Diptera: Phoridae) (Brown and Vendetti 2020; Brown 2022a). Roughly 4500 species have been described, including 1700 in the genus Megaselia, but true phorid richness could prove to be ten times greater (Brown and Hartop 2017; Brown 2022b; Brown 2022a). Phorids play many ecological roles, from fungivores and scavengers to herbivores, predators (Brown and Hartop 2017), and especially parasitoids (Brown 2022b). They are also extremely common and abundant in most parts of the world and have even been collected north of the Arctic circle (Disney 2013). Because they are species-rich, ecologically diverse, and ubiquitous, phorids make excellent model organisms for studying disturbance (Brown and Hartop 2017). They’ve been used to investigate the effects of different kinds of habitat modification from wildfire (Durska 2015), and forest management (Durska 2006), to urbanization (Durska 1981; McGlynn et al. 2019). Many species are also of economic interest as they can be honeybee parasites (Core et al. 2012), important for the bio-control invasive fire ants (Chen and Morrison 2021), used in forensic entomology (AlcaineColet et al. 2015), or serve as a model organism for biological studies (Jayakumar Pallavi et al. 2023). 59 Despite their importance and great potential as study subjects, phorids have been largely overlooked in research (Brown and Hartop 2017). Phorids are generally small, inconspicuous, and hard to identify. Many groups have complex or unresolved taxonomy or have only one sex described (Hartop et al. 2015; Brown 2022b). For example, in 1999, Disney reviewed more than 1000 Megaselia pulicaria specimens, and reclassified all except a single lectotype into different species (Disney 1999). Despite this, phorids represent an excellent opportunity for further biodiversity work, especially with the advent of DNA barcoding. Molecular sequencing technologies have become cheaper and faster in recent decades (Bansal and Boucher 2019). At the same time, emerging online repositories like GenBank and the Barcode of Life Database (BOLD) make it possible to sequence, store, and compare genetic data from thousands of animal species. However, arthropods remain proportionately severely underrepresented and identified compared to vertebrates (Meiklejohn et al. 2019; Hotaling et al. 2021). Additionally, while sequencing capacity has risen, morphological and other taxonomic efforts have not kept up due to insufficient personnel with taxon-specific expertise. In 2015, fewer than half of the invertebrate taxa added to GenBank possessed species identifications, largely because they included unnamed records from BOLD (Page 2016). Nevertheless, DNA barcoding remains one of the quickest and most affordable ways to quantify the biodiversity for obscure taxa, including phorids. For example, in 2019, more than 650 molecular operational taxonomic units (mOTUs) of phorids were detected from a single Malaise trap in Uganda using rapid sequencing 60 technology (Srivathsan et al.2019). Similarly, nearly 1500 phorid barcode index numbers (BINs) were found in Costa-Rica’s Conservación Guanacaste area in 2022 (Brown 2022a). Most recently, a 2023 Indonesian study uncovered 500 genetic sequence clusters from 5034 phorid specimens in an under-surveyed national park (Chimeno et al. 2023). While barcodes do not replace morphological and other nonDNA taxonomy, they complement it as they represent a way to gain ground when dealing with dark taxa (Srivathsan et al. 2019). Urban areas, long-regarded as impoverished habitats (Hartop et al. 2015), have largely been neglected in entomological literature save for studies focused on pests and insect disease vectors. However, as cities expand, more researchers are paying attention to how this impacts insect biodiversity (McIntyre 2009), especially in recent decades (Fenoglio et al. 2020). Phorids have proven to be very interesting focal taxa, due to their complicated relationship with urbanization. Phorids are most prevalent in forests (Durska 1981), and many depend on large swathes of natural habitat. However, phorids are often common in cities. Many species are synanthropic urbanophiles, and have even been sampled at the tops of 1000 ft buildings (Brown and Hartop 2017). Some of the earliest urban phorid research began in 1981, with Durska 1981. Durska surveyed public parks, residences, and the downtown area of Warsaw for phorids, discovering higher richness and abundance away from the urban core. She documented 36 species (Durska 1981) in that city. Twenty years later Disney (2001) recorded 53 Megaselia species from a single garden in London (Hartop et al. 2015). 61 In 2009, more work began to emerge in North America as Disney and Brown reported two new phorids for the Nearctic region from a California backyard (Disney and Brown 2009). In 2015, Hartop et al. described 30 new species from a Bioscan survey of 30 backyards in Los Angeles (L.A.) (Hartop et al. 2015). Further work unearthed nearly 100 species present in L.A., and the sites closest to natural areas were the most species rich (Brown and Hartop 2017). However, McGlynn et al. 2019 found that distance from natural areas had no significant effect on richness. Instead, temperature was the most important driver of phorid diversity and abundance. Further research into phorids can help us better understand how they respond to urbanization and what environmental variables drive these patterns. As phorids are an important component of urban biotic communities, understanding their assemblages can be useful for ecosystem services management and assessment. 3.2. OBJECTIVES The main goals of this project were to I) Inventory the phorids of an urban ecosystem in British Columbia’s central interior. II) Compare phorid abundance and richness between land use types: greenbelt, residential, industrial, and edge sites. III) Model how underlying factors like temperature and land use type impact phorid abundance and richness. IV) Compare these findings with other phorid records from Global Biodiversity Information Facility (GBIF) and BOLD for British Columbia. 62 As other studies have found that phorids appear to be more abundant and diverse near or in natural areas, I expected to collect fewer species and individuals in industrial areas. 3.3. METHODS 3.3.1. Study sites This project included 30 sites that were sampled in 2022, and 12 sites that were sampled in the 2015 study described in Chapter 2 Three of the 2015 sites were resampled in 2022: Moore’s Meadows, Rolling Mix Concrete, and Connaught Hill. The 2022 sites included eight residential sites, two edge sites (residences next to forests), ten industrial sites, and ten greenbelt sites. Like the 2015 sites, residences were homes with gardens in residential-zoned areas, industrial sites were businesses in industrial-zoned areas, and greenbelt sites were forested parks within city limits. The edge sites included student housing and a private residence that both bordered immediately adjacent forests. Residential and edge sites were confirmed with owners to be pesticide free. Greenbelt sites were unmanaged within their forested canopy, and most industrial sites were mown infrequently and were unlikely to be treated with pesticides. Of the 2022 sites, G7 (Connaught Hill) involved sampling along a cliff, so one side of the cups was frequently buried. Given this, and the low trapping yields, abundance data from G7 was excluded. Site I3 (Overhang) also had all traps and temperature sensors removed after only the second collection for that site, so it was 63 excluded from the diversity data. All sites were within an 8km radius of each other, spread out over a 200 km2 area. Figure 3.1. Map of 2022 study sites. on a satellite map of Prince George from Google Earth. Point color indicates land use types. Sites (R = Residential (R1 and R3 are edge sites), G = Greenbelt, I = Industrial) are shown in the table beside. Site R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 G1 G2 G3 G4 G5 G6 G7 G8 G9 G10 I1 I2 I3 I4 I5 I6 I7 I8 I9 I10 Latitude Longitude 53.89244 -122.817 53.88611 -122.782 53.87786 -122.778 53.86733 -122.764 53.85647 -122.766 53.90983 -122.773 53.92103 -122.768 53.91319 -122.79 53.91119 -122.798 53.91808 -122.812 53.88542 -122.821 53.87147 -122.774 53.86178 -122.787 53.85167 -122.782 53.90039 -122.807 53.90528 -122.72 53.91203 -122.744 53.91669 -122.791 53.92119 -122.799 53.93692 -122.821 53.86608 -122.785 53.83925 -122.726 53.89783 -122.771 53.87742 -122.737 53.90628 -122.665 53.91314 -122.73 53.92622 -122.694 53.96736 -122.765 53.94181 -122.821 53.90408 -122.798 64 3.3.2. Sampling For 2015 collection methods, see Chapter 2. In 2022, I collected phorids using three covered pitfall traps set in two meter transects at every site. Fewer traps were used in 2022 to ensure trap catch yields would be manageable. I filled the traps with 50% propylene glycol and began sampling in late-May for a single site (R1: UNBC Residence) and set traps in early-June for all other sites. I emptied the traps roughly every two weeks until mid-September. Each site was also equipped with a temperature sensor (Kestrel or Convergence Instruments THM sensor) which were placed in sealed cups buried in the same manner as the traps and which took readings every 30 minutes. Prior to use in the field, all of the sensors were stored together in a box for a week and their mean temperatures were compared to ensure sensor brand did not impact temperature accuracy (Wilcoxon test p = 0.26, r = 0.22). I excluded 17 out of 190 temperature records due to sensors being disturbed, batteries dying, or erratic measurements. From these data, I took the mean, minimum, and maximum temperature across all readings for a given sampling period THM sensors record the minimum, maximum and mean temperature within a 30minute period while Kestrels show only one temperature value. I also measured temperature variation as temperature standard deviation for a sampling period, divided by the number of readings for each period. 65 3.3.3. Sorting and DNA sequencing I counted all specimens that amounted to more than half a fly – i.e. if they included at least a thorax with a head or abdomen and with wings and most legs attached for a total of 3730 flies collected in 2022. I compiled traits from phorid keys such as Borgmeier’s 1964 Revision of the North American Phorid Flies Part II: The Species of the Genus Megaselia, Subgenus Aphiochaeta (Diptera, Phoridae), McAlpine’s 1987 Manual of Nearctic Diptera Volume 2, and Disney’s 2013 article: An unusually rich scuttle fly fauna (Diptera, Phoridae) from North of the Arctic Circle in the Kola Peninsula, N. W. Russia. Based on these traits, and other observed differences, I sorted my phorids into different morphospecies, separated by sex. Because of the large volume of specimens, I mainly sorted only a single pitfall trap from each site/sampling period from June to September 11th. Due to time constraints, and the limited number of samples that could be selected for sequencing, I counted, but did not sort or barcode additional samples taken on May 28th, June 10th, and September18th. I sorted additional traps for particularly productive sites or sampling periods, or to find more complete specimens for certain morphospecies. I sent 380 specimens to the Centre for Biodiversity Genomics at Guelph, based upon differential morphology, for sequencing of the DNA barcode region of the CO1 gene. I made an effort to represent every morphospecies present in my samples in relation to each land use type and, as much as possible, for each site. When combined with the 2015 DNA barcode data, this amounted to 590 successfully 66 sequenced specimens, of which 530 were barcode compliant (at least 500 bp long with less than 1% ambiguous pairs (Bold Systems 2014). 3.3.4. mOTU count Using the BOLD cluster sequence function, I grouped my flies into different molecular operational taxonomic units (mOTUs) to act as proxies for species identifications. I excluded all sequences less than 400 bp long, or that included stop codons or contaminants, leaving 582 sequences. The sequences were aligned with the BOLD Aligner and distances were calculated using the pairwise deletion method. This method converts genetic sequences to peptide sequences and arranges them using a Hidden Markov Model of the COI protein. It then converts the sequences them back into nucleotides, and those with less than 2.2% divergence from each other are grouped together with single linkage clustering. Those with 4.4% divergence or less may initially be included with their nearest neighbour, but further refinement with Markov clustering can either lump or split these sequences based on variation patterns. Several possibilities are considered and scored with a Silhouette index, where the highest scoring outcomes are kept (Ratnasingham and Hebert 2013). I built species accumulation curves for all sites, and each land use type using the method “rarefaction” from the specaccum function in vegan (R). This method randomly subsamples a given number of individuals for each site to account for differences in numbers of individuals between collections (Hurlbert 1971; Oksanen et al. 2015). It repeats this process (permutations = 10000) to generate mean numbers 67 of species found per given number of individuals and their standard deviation (Oksanen et al. 2015). I combined 2022 and 2015 sites that were resampled since they were likely to contain similar species assemblages across both years. Rarefying species richness works best when samples are random, complete, and of equal sampling effort (Hurlbert 1971). This was not the case for my data as resampled sites had the greatest sampling efforts, 2022 sites were sampled for roughly three weeks longer than 2015 sites, and different sites had different numbers of specimens sequenced. However, rarefaction is commonly used in arthropod studies to account for unequal sampling (Buddle et al. 2005). True numbers of species were estimated using the estaccumR and specpool functions from the vegan R package. 3.3.5. Abundance modelling No temperature data were collected for 2015, so only 2022 specimens were considered in these models. Of the 570 trap collections, 16 were removed due to trap disturbance or the loss of contents, and all collections from G7 were excluded due to the landscape interfering with trap efficacy. Data from September 18 th was included in models, but not in biodiversity data. Fly abundance was measured as the number of flies caught per trap, per day, by site, to account for slightly different numbers of sampling days and numbers of undisturbed traps between collections. Using the lme4 and lmerTest packages in R, I built nine mixed effects models that predicted abundance as a function of mean temperature, temperature variation, and land use type. Minimum and maximum temperature were excluded as they 68 correlated strongly with mean temperature. Sites G7, and R2 were also excluded from the data, the latter due to extremely high abundance values which may have resulted from a garden, several fruit trees, and large compost pile at the site. Collection date and site were included as random intercepts in all models. For simplicity, collection dates were combined into seven sampling periods. Temperature variables were added to models as second-degree polynomial terms as I expected abundance to be low at more extreme mean temperatures, and peak at moderate temperatures. Temperature conditions varied between land use types with greenbelt sites being the coolest and industrial sites the warmest. I expected flies in different land use types to have different temperature preferences (eg. warmer industrial sites should favour flies that can tolerate hotter temperatures), and abundances also varied greatly between land use types (highest in residential, lowest in industrial) so I modelled land use type as an intercept, and an interaction term (with and without intercepts) in different models. I evaluated model quality and assumptions with the performance package in R, including the residuals vs fitted, homogeneity of variances, normality of residuals and random effects plots, and testing the heteroscedasticity of the data using the function check_heteroscedasticity. For model quality, the performance package includes indices such as R2, intraclass correlation coefficients, weights for Aikake and Bayes Information Criterion, root mean squared error, and residual standard error. It also features an overall score calculated by rescaling all indices between zero and one, then averaging their scores for each model. I then tested the 69 significance of different terms using likelihood ratio tests on nested models fitted using restricted maximum likelihood. Table 3.1. Seven sampling periods used as random intercepts in mixed models. Sampling Period Dates 1 2022-06-19 2022-06-26 2 2022-07-03 2022-07-10 3 2022-07-17 2022-07-24 4 2022-07-31 2022-08-07 5 2022-08-12 2022-08-13 6 2022-08-28 2022-09-04 7 2022-09-11 2022-09-18 Table 3.2. Model equations for nine mixed effects models (Abun = Abundance). Land use type is a 4-way factorial term. In some models, land use type is an interaction term, an intercept, or excluded. Second degree polynomial terms for temperature variables are referred to as poly(temperature variable). Models Abun = land use type * poly (mean temperature) (interaction and intercepts) Abun = land use type * poly (mean temperature) (interaction only) Abun = land use type + poly (mean temperature) (no interaction) Abun = land use type * poly (temperature variation) (interaction only) Abun = land use type + poly(temperature variation) (no interaction) Null model with random effects only Abun = poly (mean temperature) Abun = land use type Abun = poly (temperature variation) 70 3.3.6. mOTU richness by land use type I compared the richness of mOTUs by land use type using Kruskal Wallis and Dunn tests. Due to differences in sampling methods between years, I kept comparison of flies from 2022 and 2015 separate. I also explored the habitat preference of different mOTUs using a Venn diagram built with the ggvenn package and rpb index from the indicspecies package (Table 5.2.2.). 3.3.7. Identifications and BINs BOLD clusters sequences from its overall database using neighbour joining trees. Terminal clusters are assigned barcode index numbers or BINs which, like mOTUs, can act as species proxies but include sequences from public and private data from BOLD, while mOTUs in this case, were grouped using data from only this project. BINs may be split or rearranged when new 500 bp sequences are added and help refine relationships between specimens (BOLD Systems 2011). Sequences can be compared on BOLD using the complete BOLD database which includes sequences at least 500 base pairs (bp) long or using the BOLD species database which excludes sequences without an associated species identification (BOLD Systems 2011). I wanted to know what potential species names or BINs my mOTUs belonged to. Using the Batch ID engine from BOLD, I compared sequences from my project with those from existing BINs and species names recorded in the BOLD database. I used sequences at least 400 bp long, with 400 bp of overlap. I first searched the database for matches with species identifications. In R, I filtered those results to include only 100%, 99%, or 98.5% matches, and grouped them by mOTU. I chose conservative divergence thresholds (≤1.5%) because my minimum 71 sequence length was short and broader matches tended to result in more species names for a single mOTU. A single specimen from this project could match with more than a hundred records from the BOLD, so I averaged the percent match at each filtration level for each unique species identification. Most species identifications also had corresponding BINs. For those that did not, I repeated my search using the full BOLD database, and filtered results using a 98.5% match threshold. For results that matched neither species nor BIN, I would search again without a match percent threshold, or BLAST the genetic sequence on GenBank, which helped indicate specimen genus (Table 5.2.1.). 3.3.8. Phorid records in British Columbia from GBIF and BOLD To compare these results with other species and BINs already recorded in British Columbia (BC), I extracted public records from GBIF and BOLD. On September 18th 2023, I downloaded 102302 phorid records from GBIF. I drew a polygon around British Columbia, and used the TaxonKey Phoridae, including records that were preserved and material specimens. Assuming that coordinates were recorded in the World Geodetic System 1984 (WGS 84), I assigned occurrence records to different ecoregions using the the “bcmaps” package. I excluded all observations recorded outside BC (52108 observations), and without ecoregion data (18157 observations). For sites that had no province recorded, but did have an ecoregion, I checked the locality data to confirm their location, and excluding 57 blank records. I grouped the resulting 31980 records by species. 72 On September 25th 2023, I downloaded 12418 records from the BOLD public database using the search terms “Phoridae British Columbia”. I excluded four records without coordinates. I plotted occurrence records in different ecoprovinces using the ggplot2, sf, and the bcmaps package in R. I compared the number of species and BIN records between GBIF, BOLD and this project. Using data from the Sub-Boreal Interior ecoprovince, I repeated the procedure. 73 3.4. RESULTS 3.4.1. Data Quality Figure 3.2. shows the number of flies collected, sorted, and barcoded by site for 2022. Of the 3730 flies collected, 58% were sorted morphologically (64% for selected dates), and 9.4% were genetically sequenced (10.3% for selected dates). Figure 3.2. Abundance of flies by site for all dates (red), selected dates (2022-06-19 to 202209-11) which excluded early and late collections so that sites were comparable in their sampling effort (blue). The green and white bars show the number of flies, by site, that were morphologically sorted and the number that were DNA-barcoded, respectively. 74 Figure 3.3. shows the number of morphospecies, specimens barcoded, and mOTUs detected by site. On average, mOTUs estimated only 55% of the richness values per site compared to morphological sorting. Morphospecies likely represent an overestimate as male and female phorids are dimorphic, so a single species could count as two morphospecies. The number of mOTUs likely represents the most conservative estimate as only a proportion of the flies sampled could be sequenced (roughly seven specimens for every ten morphospecies estimated by site). Of the 132 morphospecies assessed to be in Prince George in this study, only seven had no recovered barcodes due to failed sequencing at the service provider. 75 Figure 3.3. The number of all, female, and male morphospecies estimated by site, the number of specimens barcoded, and the number of mOTUs actually detected for greenbelt, residential, edge, and industrial sites land use types. For flies collected between 2022-06-19 and 2022-09-11, when sampling methods were consistent, greenbelt, residential, and industrial sites did not differ significantly from each other in the number of samples sorted (ANOVA, df = 3, F = 2, p = 0.14), the proportion of flies sorted (ANOVA, df = 2,24, F = 0.37, p = 0.70), and the proportion of flies sequenced (ANOVA, df = 2,24, F = 1.47, p = 0.25), or the ratio of morphospecies identified per site to the number of specimens barcoded (ANOVA, 76 df = 2,24, F = 0.51, p = 0.61), suggesting a similar sampling effort between land use types (Table 3.3). Table 3.3. The number of flies counted and sorted for each land use type. The mean ratio of morphospecies estimated (MSE) per specimens barcode, and the mean ratio of morphospecies estimated to the number of mOTUs detected. Land use type Gb Indus Res Edge Fly count for all dates 1218 375 1808 329 Number of flies sorted 749 209 1069 153 Mean ratio of MSE to specimens barcoded per site 1:0.67 1:0.74 1:0.72 1:1.36 Mean ratio of MSE to unique mOTUs detected 1:0.53 1:0.61 1:0.52 1:1.11 Flies from 2015 were sampled from fewer sites (n=12) and were not sorted to morphospecies. Most flies caught during the selected dates from chapter 2 (17 July, 24 July, 6 August, and 28 August 2015) were sent to BOLD. In 2015, more than 100 flies were caught at the Connaught Hill site, most of those belonging to Megaselia arcticae. 3.4.2. Richness estimates Of the 590 flies sequenced, I detected 99 mOTUs (Figure 3.4.). Estimates (Bootstrap and Jackknife estimates respectively) suggest there could be an additional 18 ± 6 to 61 mOTUs that have not yet been detected. However, these estimates assume equal sampling effort between sites and years, which is not the case for this data. Greenbelt sites had the highest richness, followed by residential and industrial sites with 67, 45, and 28 mOTUs. Chao 1 analysis gave maximum estimates of 188 ± 60, 79, and 63 mOTUs, respectively (Table 3.4.). Edge sites were not included in these estimates as only two sites were sampled, although they proportionally were overrepresented in the barcoding data. 77 Figure 3.4. The accumulation curve of mOTUs detected from the 582 sequenced specimens across 40 sites, over two sampling years (2015 and 2022). Using the specaccum (method=rarefaction) package from vegan in R, individuals were added randomly in 10000 permutations of the data to find the mean species accumulation curve. Dots represent the mean number of mOTUs found per number of sites sampled (± one standard deviation in blue). A total 99 mOTUs were detected across the sites and sampling years. 78 Figure 3.5. The accumulation curves of mOTUs for Prince George where the black line represents the mean mOTU richness by site, red the richness estimated by the Chao 1 index, and blue the richness estimated by the Abundance Coverage index from estimateR in vegan. 79 Figure 3.6. The accumulation curves of mOTUs detected in greenbelt (green), residential (blue), and industrial (gray) land use types, using the specaccum (method=rarefaction) package from vegan in R. Individuals were added randomly in 10000 permutations of the data to find the mean species accumulation curve. Dots represent the mean number of mOTUs found per number of sites sampled (± one standard deviation which is the shaded area) Table 3.4. Estimates of mOTU richness for the regional species pool and each land use type, along with their ± one standard error for the Chao 2 index, Jackknife 1, Jackknife2, Bootstrap, Chao 1, and Abundance Coverage estimators. Land use type All Greenbelt Residential Industrial Number of sites 40 12 11 12 mOTU richness 99 67 45 28 Chao 2 Jack 1 142 ± 18 188 ± 60 71 ± 14 57 ± 21 139 ± 10 106 ± 15 67 ± 10 43 ± 7 Jack 2 160 137 79 53 Bootstrap 117 ± 6 83 ± 6 55 ± 5 34 ± 4 Chao 1 122 110 60 63 ACE 133 129 76 45 80 3.4.3. Fly abundance modelling Phorid abundance (the number of Phoridae caught per trap per day) peaked in late July to mid August, when mean temperatures were at their highest for the summer (Figures 3.6. and 3.7.). Despite industrial sites having the highest mean temperatures, they also had the lowest trap yields, with many traps catching almost no flies throughout most of the summer. Residential, edge, and greenbelt sites were increasingly shaded and thus cooler, in that order (Figure 3.7.), but also had much higher trap yields (Figure 3.8.). The other land use types with more vegetation cover (especially tall vegetation such as trees and shrubs), generally supported greater numbers of phorids than industrial sites. Phorid abundance peaked at moderate mean temperatures – near 15 °C and decreased once temperatures exceeded 18 °C .(Figure 3.10.).Temperature variation was also much higher at industrial sites and lowest at greenbelt sites, indicating industrial site flies also had to cope with greater fluctuations in temperature (Figure 3.9.). 81 Figure 3.6. The mean number of flies caught per trap per day by collection date during the summer of 2022. Most trapping periods were two-weeks in length. Residential and greenbelt sites had the highest trap yields while industrial sites had the lowest. 82 Figure 3.7. Mean temperature (°C) by date for the summer of 2022. Industrial sites were consistently the warmest while greenbelt sites were the coolest and residential and edge sites were intermediate. The temperature peaked in August for most sites. 83 Figure 3.8. Mean fly trapping rate over the summer of 2022 by land use type. The number of flies caught per trap, per day per site was averaged over the course of the summer – excluding I3 (too few collections retrieved) and G7 (trapping rates potentially affected by trap position). Data from outside selected dates (May 28th, June 10th, and September18th) were included as their trapping rates were not very different than those from collection periods directly before, or after them (assessed visually using figure 3.6.). 84 Figure 3.9. Temperature variation (temperature standard deviation/number of temperature readings) by collection dates. Industrial sites had the highest variation and greenbelt sites the lowest. Variation was highest in August, but less stable on a weekly basis than mean temperature. Of the nine models tested for their ability to predict fly abundance, the full model including mean temperature and land use type performed best (Table 3.5.). It scored the highest performance score from the performance package and all its predictors proved significant in maximum likelihood ratio tests. The model suffers from two main weaknesses. The first is that it contains many terms, including polynomial terms which can be complex to interpret and can be highly correlated (Table 3.7.). This makes it difficult to separate model variation among different terms as those terms represent related, and non-linear relationships. The second issue is that because greenbelt sites never reached temperatures extreme enough to reduce fly abundance, the slope of the greenbelt model predicted exponential abundance 85 growth with increasing temperatures rather than declining with extreme heat (Table 3.12.). When land use type was included as an intercept term rather than a slope/interaction term (Abun = land use type + mean temperature (no interaction)), the model did not predict increasing abundance with increasing temperature for greenbelt sites and contained fewer terms (Figure 3.10.). However, it was not as good at predicting abundance values and its scores were lower according to performance’s index (Table 3.5.). Nevertheless, mean temperature, and land use type were both significant predictors of fly abundance, while temperature variation was not (Table 3.6.). Table 3.5. Indices ranking the performance of nine fly abundance models, built using the compare_performance function from the performance package in R. Indices include conditional R2 (R2 cond.), marginal R2 (R2 marg.), intraclass correlation coefficient (ICC), root mean squared error (RMSE), residual standard error (RSE), corrected Aikake Information Criterion weight (AICc wt), Bayesian Information Criterion weight (BIC wt), and the performance score (Score). Model Abun = land use type * mean temperature (interaction and intercepts) Abun = land use type * mean temperature (interaction only) Abun = land use type + mean temperature (no interaction) Abun = land use type * temperature variation (interaction only) Abun = land use type + temperature variation (no interaction) Null model with random effects only Abun = mean temperature Abun = land use type Abun = temperature variation R2 R2 cond. marg. ICC RMSE RSE AICc wt. BIC wt. Score 0.64 0.19 0.56 0.27 0.30 0.69 0.00 0.81 0.64 0.08 0.61 0.27 0.30 0.16 0.00 0.59 0.62 0.14 0.55 0.28 0.32 0.00 0.00 0.34 0.60 0.05 0.57 0.28 0.32 0.00 0.00 0.27 0.59 0.15 0.52 0.29 0.32 0.06 0.00 0.24 0.56 0.00 0.56 0.29 0.32 0.01 0.85 0.21 0.58 0.58 0.56 0.03 0.12 0.03 0.56 0.52 0.54 0.29 0.29 0.30 0.32 0.32 0.33 0.03 0.04 0.00 0.10 0.03 0.02 0.19 0.14 0.05 86 Table 3.6. The results of likelihood ratio tests for nested sets of models fitted using restricted maximum likelihood. (Degrees of freedom= d.f., p-value = p, Abun = Abundance) Model 1 Null Null Model 2 Abun = land use type Abun = poly (mean temperature) Null Abun = poly (temperature variation) Abun = land use type + poly (mean temperature) Abun = land use type * poly (mean temperature) (interaction only) Abun = land use type * poly (mean temperature) (interaction and intercepts) Abun = land use type * poly (mean temperature) (interaction and intercepts) Abun = land use type*poly (temperature variation) (interaction only) Abun = land use type*poly (temperature variation) (interaction and intercepts) Abun = land use type Abun = land use type Abun = land use type + poly (mean temperature) Abun = land use type * poly (mean temperature) (interaction only) Abun = land use type Abun = land use type*poly (temperature variation) (interaction only) ߯ଶ 8.63 5.90 d.f. 3 2 p 0.03* 0.05* 2.31 2 0.31 5.62 2 0.06 14.18 5 0.01* 18.78 6 0.0046* 10.22 3 0.02* 3.77 5 0.58 7.01 3 0.07 87 Table 3.7. The variable estimates for the full model of mean temperature. (Mean Temperature = Mean. Temp., Mean Temperature 2nd degree polynomial term =Mean. Temp2) Variable Intercept Gb Indus Res Mean Temp. Mean Temp.2 Gb * Mean Temp Indus * Mean Temp Res * Mean Temp Gb * Mean Temp2 Indus * Mean Temp2 Res * Mean Temp2 Estimate 0.64 0.08 -0.47 -0.01 -0.25 -0.24 0.91 0.26 0.34 0.50 0.22 0.00 s. e. 0.25 0.27 0.27 0.27 0.24 0.23 0.28 0.25 0.24 0.25 0.23 0.25 T 2.62 0.30 -1.73 -0.03 -1.07 -1.04 3.26 1.04 1.44 2.02 0.95 -0.02 Figure 3.10. Mean fly trapping rate by mean temperature with curves representing the slope of the model with mean temperature and land use type included as an intercept. Each land use type had the same slope. Edge slope (orange) hidden by greenbelt slope (green). 88 Figure 3.11. Mean fly trapping rate by mean temperature with slopes from the full model. Each land use type has its own slope and intercepts. 3.4.4. mOTU richness by land use type mOTUs richness was significantly lower in industrial sites compared to other land use types in 2022 (Table 3.8.). Residential sites in 2015 had far fewer flies and mOTUs compared to residential sites in 2022, and significantly fewer flies compared to 2015 greenbelt sites. Edge sites had the highest number of mOTUs on average, while industrial sites generally had low richness regardless of the year of sampling (Figure 3.12.). However, edge sites had more specimens sent for barcoding per morphospecies estimated than other land use types, and fewer sites were sampled. So, while they may have high richness, flies from edges sites are overrepresented in the data and variation in richness may not have been captured by only sampling two sites. Likewise, 2022 samples were not sequenced equally or randomly. Specimens were selected maximize the number of morphospecies represented per land use 89 type, and site as much as possible. While greenbelt, residential, and industrial sites were comparable in the proportion of specimens sequenced per number of morphospecies identified, some sites were better represented in the data than others, and some mOTUs, especially cryptic species, may have been missed. Table 3.8. Comparison of mOTU richness by land use type and for 2015 and 2022 at selected dates with Kruskal-Wallis and Dunn tests. Year 2015 2015 2015 2015 2022 2022 2022 2022 2022 2022 2022 Test Kruskal-Wallis Dunn Test Dunn Test Dunn Test Kruskal-Wallis Dunn Test Dunn Test Dunn Test Dunn Test Dunn Test Dunn Test Comparison All Gb-Indus Gb-Res Indus-Res All Edg-Gb Edg-Indus Gb-Indus Edg-Res Gb-Res Indus-Res Chi-sq 8.55 8.55 8.55 8.55 10.05 10.05 10.05 10.05 10.05 10.05 10.05 Z df 2 1.68 2.91 1.23 3 1.10 2.37 2.10 0.73 -0.58 -2.61 p-value 0.01 0.05 0.00 0.11 0.02 0.14 0.01 0.02 0.23 0.28 0.00 adj. pvalue 0.14 0.01* 0.33 0.81 0.05* 0.11 1.00 1.00 0.03* 90 Figure 3.12. The number of mOTU detected for each site where color represents land use type and shape represents year. Industrial sites generally had low richness, edge sites, high richness and greenbelt and residential sites had high variability. Site G7 and I3 were excluded. 91 Greenbelt sites had the most unique mOTUs (31) – i.e., mOTUs found nowhere else – followed by residential (11), industrial (6), and edge (3) sites (Figure 3.13.). Some edge site unique mOTUs may have resulted from those sites status as effectively a combination of greenbelt and residential habitats – and only a few sites were sampled, giving a smaller species pool to draw from. mOTU overlap may increase with further sequencing. Only five mOTUs were present at all land use types, potentially indicating a high number of habitat specialists in greenbelt sites, more than forty flies found in two to three land use types, and limited number of eurytopic flies in the overall species pool when considering all habitat types. However, only mOTU-8 was significantly associated with greenbelt habitat (rpb= 0.59, p = 0.01). Figure 3.13. Venn diagram of mOTUs unique to and shared by land use types using sequenced flies from 2022 and 2015. Shade indicates the number of mOTUs found in each category. The figure was made using the ggvenn package in R. 92 3.4.5. Provincial data from GBIF and BOLD databases By comparing gene sequences from Prince George with records from the Barcode of Life and GenBank databases, I detected 80 BINs and 31 species matches. BOLD assigned an additional 16 BINs to Prince George specimens that did not appear in the Batch ID engine search results, likely because of the conservative search criteria. Additionally, one BIN that did not appear in search results had most of its members originating from this project or matching specimens from private data that was inaccessible at the time of this writing. Two mOTUs matched with more than one species name, and a few species names were associated in common for several different mOTUs. This is not unexpected given the current state of phorid taxonomy (Brown 2022b), and the potential for misidentifications in database records (Pentinsaari et al. 2020). When more than one species name appeared for a single mOTU, the names were excluded. Prince George specimens matched ten different genera, with the vast majority of mOTUs belonging to the hyperdiverse genus Megaselia. When comparing with online databases, GBIF listed 103 species names (15 genera) and 2 BINs, while public BOLD records listed 81 species names, (11 genera), and 520 BINs. In both cases, most observations originated in southern British Columbia, concentrated around Vancouver and Victoria, leaving much of the province unexplored. Observations from GBIF and BOLD had similar rates of species identifications: 22.6 % and 25.2 %. However, 99.6 of BOLD observations had associated BINs compared to 0.9 % of observations from GBIF. When considering the Sub-Boreal Interior ecoprovince, both GBIF and BOLD listed fewer 93 observations than I found in my research – recording 12 species and 2 BINs, and 8 species and 37 BINs, respectively. Table 3.9. The number of mOTUs from PG belonging to each genus. Identifications were made from morphological sorting, as well as matches on GenBank and BOLD. Genus Aenigmatias Beckerina Diplonevra Gymnophora Lecanocerus Megaselia Metopina-group Metopina Phora Pseudoaceton Puliciphora Number of mOTUs 1 1 1 1 1 86 1 1 2 1 1 94 Figure 3.14. Map of 12414 phorid observations downloaded from BOLD on September 24th, 2022.This map was built using the BCmaps and sf package. 95 Figure 3.15. The number of species and barcode index numbers (BINs) recorded in GBIF, and the Barcode of Life public database from September 2023 for the entire province of British Columbia and BC’s Sub-Boreal Interior ecoprovince specifically. Prince George refers to flies collected from this project. 96 Figure 3.16. Venn diagram of species unique to and shared between GBIF, BOLD, and the Prince George project. 3.5. DISCUSSION Prince George hosts a high diversity of phorids including 99 distinct mOTUs, with potentially another 61 mOTUs yet to be found. In tropical biodiversity hotspots, researchers have caught as many as 650 mOTUs in a single trap (Srivathsan et al. 2019). This project resembles the 2015 Los Angeles Bioscan project in scope, which also sampled 30 sites and uncovered 99 species (Brown and Hartop 2017). Yet the LA study took place at a lower latitude (~34 N) over a larger area, and caught roughly eleven times the amount of phorids using Malaise traps (Brown and Hartop 2017). Despite being less urbanized and arid than LA, its still surprising Prince George hosts similar levels of diversity. However, given that more than 60 species 97 were found in Russia’s far north (~67°N) in one study (Disney 2013), it seems that phorid diversity can be high, even at moderate latitudes. Industrial sites generally had poor phorid abundance and diversity, suggesting heavy urbanization does not favour phorids. Temperature is a likely explanatory factor, as industrial sites were significantly warmer and more variable than greenbelt sites, and abundance peaked at moderate temperatures typical of greenbelt sites. McGlynn et al. 2019 found that temperature had a stronger influence on phorid diversity and abundance than land cover. However, abundance may have been more affected by land use type for industrial sites as the abundance remained low across a range of temperatures, while other land use types had distinct peaks. Phorids occupy a wide range of niches (Brown and Hartop 2017), so it is difficult to determine what factors may affect phorid abundance. However, resources like gardens, composts, fungi, and decomposing vegetation may better provide phorids with certain food resources or hosts (Durska 2015; Hartop et al. 2018). The residential site with the highest abundance (R2) also had a massive compost pile, garden, and decomposing apples at the site, which might have contributed to the high trap rates. This may also be part of the reason industrial sites had such low richness. Richness for the 2022 flies was significantly higher in residential and edge sites compared to industrial sites. I7, the industrial site with the highest richness and abundance, also had very tall, unmown vegetation suggesting good vegetation cover could be favourable to phorids in otherwise less-suitable settings. This is consistent with Durska 1981, who found the highest phorid richness in Warsaw parks compared 98 to the city center, and described phorids as most common in forests, often occurring in wet soil or decaying organic matter. Greenbelts sites also had the most unique mOTUs, suggesting many phorids might be forest specialists. Notably, the 2015 residential sites had much lower abundance and significantly lower richness compared to greenbelt sites. 2015 sites had a shorter sampling period, but low richness may also be due to site specific differences, or different environmental conditions during the summer of 2015. Temporal variation should be further explored. Prince George has a rich assemblage of phorid species, and given that many species have unknown life histories (Hartop et al. 2018; Brown 2022b), their full contribution to functional diversity goes unappreciated. Because they are species rich, diverse in their niches, and important colonizers to habitats recovering from various types of disturbance, they could also be valuable indicators of environmental health (Disney and Durska 2008). Given that one of the main differences between industrial and other land use types is vegetation cover, phorid diversity could potentially be improved at industrial sites by growing taller vegetation such as long grass or trees. Shading would also help reduce and stabilize temperature (Shiflett et al. 2017). While GBIF and BOLD both represent extensive repositories of taxonomic records, they are far from comprehensive (Meiklejohn et al. 2019; Garcia-Rosello et al. 2023). GBIF records contained many scientific names, but most phorids in their database lacked detail to the species level. This is especially difficult as even museum specimens may be misidentified or belong to groups with unresolved 99 taxonomy (Disney 2002). Public records from the Barcode of life database recorded fewer scientific names, but roughly five times more identifications than GBIF in the form of Barcode Index Numbers. BINs provide a way to name and track specimens from cryptic or taxonomically complex taxa. However, BINs do not necessarily represent accurate delimitations of species, and require morphological evidence to support grouping (Young et al. 2019; Prieto et al. 2021). Most identifications from BOLD (species and BINs) matched with what had previously been recorded for the province of BC on BOLD and GBIF. 6 BINs that appeared unique to the Prince George dataset had previously been recorded near or in BC when private records were checked. mOTUs that did not strongly match sequences from the BOLD database may belong to BINs with large intraspecific variation, require higher quality sequences, or more specimens, and more public records. It is currently not possible to know if any mOTUs from this project represent new species (either undescribed species or species that weren’t previously recorded in BC), given the lack of available data to compare specimens against (Page 2016), and the challenges of morphological sorting for this group. However, this project has helped fill in one of the geographic gaps in phorid records for the province, as BINs and species names from this project outnumbered those previously recorded in public data for BOLD and GBIF for the Sub-Boreal Interior ecoprovince. Likewise, vouchered Prince George specimens and sequences have helped quantify potential biodiversity in the area and can serve as reference for future projects. 100 Despite recent efforts which have uncovered unexpectedly high phorid diversity in many parts of the world (Namaki-Khameneh et al. 2021; Brown 2022a), much work remains to be done for this taxon, both in building comprehensive species records, and in untangling their taxonomy (Brown 2022a; Chimeno et al. 2023). Future work should prioritize collecting phorids in under-sampled areas, collecting DNA from holotypes and type specimens for better reference sequences (Prosser et al. 2016), and revising phorid taxonomy from both old and newly collected specimens using both morphological and genetic data (Brown 2022a; Hartop et al. 2022). 3.6. REFERENCES Alcaine-Colet A, Wotton K, Jimenez-Guri E. 2015. Rearing the scuttle fly Megaselia scalaris (Diptera: Phoridae) on industrial compounds: Implications on size and lifespan. PeerJ. 3:e1085. doi:10.7717/peerj.1085. Bansal V, Boucher C. 2019. Sequencing technologies and analyses: Where have we been and where are we going? iScience. 18:37–41. doi:10.1016/j.isci.2019.06.035. BOLD Systems. 2011. The Barcode of Life Systems Handbook. Bold Systems. 2014. Record List. BOLD Handbook. [accessed 2024 Feb 26]. https://v3.boldsystems.org/index.php/resources/handbook?chapter=6_managingdata.html&s ection=record_list. Brown B. 2022a. Some remarkably common, but undescribed, Megaselia Rondani (Diptera: Phoridae) from northwestern Costa Rica. Zootaxa. 5120:373–390. doi:10.11646/zootaxa.5120.3.4. Brown B. 2022b. Two new genera of phorid flies, Macgrathphora and Aurisetiphora, from Costa Rica (Diptera: Phoridae), with recommendations for naming new genera in the family. Zootaxa. 5115:571–581. doi:10.11646/zootaxa.5115.4.7. Brown BV, Hartop EA. 2017. Big data from tiny flies: patterns revealed from over 42,000 phorid flies (Insecta: Diptera: Phoridae) collected over one year in Los Angeles, California, USA. Urban Ecosys. 20(3):521–534. doi:10.1007/s11252-016-0612-7. Brown BV, Vendetti JE. 2020. Megaselia steptoeae (Diptera: Phoridae): Specialists on smashed snails. Biodivers. Data J. 8(Journal Article):e50943–e50943. doi:10.3897/BDJ.8.e50943. 101 Buddle CM, Beguin J, Bolduc E, Mercado A, Sackett TE, Selby RD, Varady-Szabo H, Zeran RM. 2005. The importance and use of taxon sampling curves for comparative biodiversity research with forest arthropod assemblages. Can. Entomol. 137(1):120–127. doi:10.4039/n04-040. Chen L, Morrison LW. 2021. Importation biological control of invasive fire ants with parasitoid phorid flies—Progress and prospects. Biol. Control. 154:104509. doi:10.1016/j.biocontrol.2020.104509. Chimeno C, Schmidt S, Hamid H, Narakusumo RP, Peggie D, Balke M, Cancian de Araujo B. 2023. DNA barcoding data release for the Phoridae (Insecta, Diptera) of the HalimunSalak National Park (Java, Indonesia). Biodivers. Data J. 11(Journal Article):e104942– e104942. doi:10.3897/BDJ.11.e104942. Core A, Runckel C, Ivers J, Quock C, Siapno T, DeNault S, Brown B, DeRisi J, Smith CD, Hafernik J. 2012. A new threat to honeybees, the parasitic phorid fly Apocephalus borealis. PLOS One. 7(1):e29639. doi:10.1371/journal.pone.0029639. Disney R. 2002. Revisionary notes on European Phoridae (Diptera). Bonn. Zool. Beitr. 50(4):293–304. Disney RHL. 1999. A troublesome sibling species complex of scuttle flies (Diptera: Phoridae) revisited. J. Nat. Hist. 33(8):1159–1216. doi:10.1080/002229399299987. Disney RHL. 2013. An unusually rich scuttle fly fauna (Diptera, Phoridae) from north of the Arctic Circle in the Kola Peninsula, N. W. Russia. Zookeys.(342):45–74. doi:10.3897/zookeys.342.5772. Disney RHL, Brown BV. 2009. Two additions to the list of phorid flies (Diptera: Phoridae) reported from the USA. Proc. Entomol. Soc. Wash. 111(2):539–540. doi:10.4289/0013-8797111.2.539. Disney RHL, Durska E. 2008. Conservation evaluation and the choice of faunal taxa to sample. Biodivers. Conserv. 17(3):449–451. doi:10.1007/s10531-007-9284-1. Durska E. 1981. Phoridae (Diptera) of Warsaw. Polska Akademia Nauk. Instytut Zoologii, editor. Species Composition and Origin of the Fauna of Warsaw ; pt 2 (3). 35:47–56. Durska E. 2006. Diversity of the scuttle fly (Diptera: Phoridae) communities in the plantations of moist pine forests of the Białowieża Primeval Forest and the Tuchola Forest (Poland). Biodivers. Conserv. 15(1):385–393. doi:10.1007/s10531-005-3432-2. Durska E. 2015. Effects of fire on scuttle flies (Diptera: Phoridae) in a pine forest in Poland. Entomol. Fenn. 26(4):181–193. doi:10.33338/ef.84645. Fenoglio MS, Rossetti MR, Videla M, Baselga A. 2020. Negative effects of urbanization on terrestrial arthropod communities: A meta-analysis. Baselga A, editor. Glob. Ecol. Biogeogr. 29(8):1412–1429. doi:10.1111/geb.13107. Forister ML, Pelton EM, Black SH. 2019. Declines in insect abundance and diversity: We know enough to act now. Conserv. Sci. Pract. 1(8):e80. doi:10.1111/csp2.80. 102 Garcia-Rosello E, Gonzalez-Dacosta J, Guisande C, Lobo JM. 2023. GBIF falls short of providing a representative picture of the global distribution of insects. Syst. Entomol. 48(4):489–497. doi:10.1111/syen.12589. Hartop E, Srivathsan A, Ronquist F, Meier R. 2022. Towards large-scale integrative taxonomy (LIT): Resolving the data conundrum for dark taxa. Syst. Biol. 71(6):1404–1422. doi:10.1093/sysbio/syac033. Hartop EA, Brown BV, Disney RHL. 2015. Opportunity in our ignorance: urban biodiversity study reveals 30 new species and one new Nearctic record for Megaselia (Diptera: Phoridae) in Los Angeles (California, USA). Zootaxa. 3941(4):451–484. Hartop EA, Gonzalez LA, Brown BV. 2018. Backyard biodiversity: Unraveling life histories of the new fly species discovered by the BioSCAN Project proves harder than first assumed. Journal of Negative Results. 12:1–8. Hotaling S, Kelley JL, Frandsen PB. 2021. Toward a genome sequence for every animal: Where are we now? Proc. of the Natl. Acad. Sci. U.S.A. 118(52):e2109019118. doi:10.1073/pnas.2109019118. Hurlbert SH. 1971. Nonconcept of species diversity: A critique and alternative parameters. Ecology (Durham). 52(4):577–586. doi:10.2307/1934145. Jayakumar Pallavi, Harshita Snehal, Daniela Salazar, Mrunal Hanbar, Larina Bejoy Chiramel, Khushi Alok Jha, Rakshita Sukruth Kolipakala, Sai Bhumica Lakshmi Venkatesh, Tanishka Dayanand Shetty, Navya Madhusudan, et al. 2023. A complete morphological characterization of all life stages of the phorid fly M. scalaris. bioRxiv.2023.05.18.541282. doi:10.1101/2023.05.18.541282. McGlynn TP, Meineke EK, Bahlai CA, Li E, Hartop EA, Adams BJ, Brown BV. 2019. Temperature accounts for the biodiversity of a hyperdiverse group of insects in urban Los Angeles. P. Roy. Soc. B, Biol. Sci. 286(1912):20191818. doi:10.1098/rspb.2019.1818. McIntyre N. 2009. Ecology of urban arthropods: A review and a call to action. Ann. Entomol. Soc. Am. 93:825–835. doi:10.1603/0013-8746(2000)093[0825:EOUAAR]2.0.CO;2. Meiklejohn K, Damaso N, Robertson J. 2019. Assessment of BOLD and GenBank – Their accuracy and reliability for the identification of biological materials. PLOS One. 14:e0217084. doi:10.1371/journal.pone.0217084. Namaki-Khameneh R, Khaghaninia S, L. Disney RH, Maleki-Ravasan N. 2021. The scuttle flies (Diptera: Phoridae) of Iran with the description of Mahabadphora aesthesphora as a new genus and species. Zhang F, editor. PLOS One. 16(10):e0257899–e0257899. doi:10.1371/journal.pone.0257899. Oksanen J, Blanchet FG, Kindt R, Legendre P, Minchin P, O’Hara B, Simpson G, Solymos P, Stevens H, Wagner H. 2015. Vegan: Community Ecology Package. R Package Version 221. 2:1–297. Page RDM. 2016. DNA barcoding and taxonomy: Dark taxa and dark texts. Philos. Trans. Roy. Soc. Lond. B: Biol. Sci. 371(1702):20150334–20150334. doi:10.1098/rstb.2015.0334. 103 Pentinsaari MD curation, Ratnasingham SI, Miller SEI, Hebert PDNC, Kuntner ME. 2020. BOLD and GenBank revisited – Do identification errors arise in the lab or in the sequence libraries? PLOS One. 15(4). doi:10.1371/journal.pone.0231814. Petchey O, O’Gorman E, Flynn D. 2009. A functional guide to functional diversity measures. In: Biodivers. Ecosyst. Funct. Hum. Wellbeing Ecol. Econ. Perspect. Vol. 85. Prieto C, Faynel C, Robbins R, Hausmann A. 2021. Congruence between morphologybased species and Barcode Index Numbers (BINs) in Neotropical Eumaeini (Lycaenidae). PeerJ (San Francisco, CA). 9(Journal Article):e11843–e11843. doi:10.7717/peerj.11843. Prosser SWJ, deWaard JR, Miller SE, Hebert PDN. 2016. DNA barcodes from century-old type specimens using next-generation sequencing. Mol. Ecol. Resour. 16(2):487–497. doi:10.1111/1755-0998.12474. Ratnasingham S, Hébert PDN. 2007. bold: The Barcode of Life Data System ( http://www.barcodinglife.org). Mol. Ecol. Notes. 7(3):355–364. doi:10.1111/j.14718286.2007.01678.x. Ratnasingham S, Hebert PDN. 2013. A DNA-based registry for all animal species: the barcode index number (BIN) system. Fontaneto D, editor. PLOS One. 8(7):e66213–e66213. doi:10.1371/journal.pone.0066213. Shiflett SA, Liang LL, Crum SM, Feyisa GL, Wang J, Jenerette GD. 2017. Variation in the urban vegetation, surface temperature, air temperature nexus. Sci. Total Environ. 579:495– 505. doi:10.1016/j.scitotenv.2016.11.069. Srivathsan A, Hartop E, Puniamoorthy J, Lee WT, Kutty SN, Kurina O, Meier R. 2019. Rapid, large-scale species discovery in hyperdiverse taxa using 1D MinION sequencing. BMC Biol. 17(1):96–96. doi:10.1186/s12915-019-0706-9. Young MR, Moraza ML, Ueckermann E, Heylen D, Baardsen LF, Lima-Barbero JF, Gal S, Gavish-Regev E, Gottlieb Y, Roy L, et al. 2019. Linking morphological and molecular taxonomy for the identification of poultry house, soil, and nest dwelling mites in the Western Palearctic. Sci. Rep. 9(1):5784. doi:10.1038/s41598-019-41958-9. 104 Chapter four SUMMARY Despite its small geographic size, Prince George harbours a rich assemblage of ground arthropods, across several different land use types. Industrial, residential, and greenbelt areas host similar levels of taxonomic and functional diversity. If they can foster high ecological diversity, highly urbanized spaces like industrial parks and yards should be managed for their conservation value. Future studies could test ways of improving arthropod diversity such as reduced mowing, diversifying lawn plants, and increasing green space size and continuity. Simultaneously, forested areas should be protected because they harbour unique sets of species and may better support natural enemies. Many studies report the negative effects of urbanization. Given the limited scope of this project, further research should look at the long-term effects of urban development, as well as the potential for yearly variation in arthropod diversity. Nevertheless, these results suggest that cities can be biodiversity friendly and support a myriad of ecosystem services. Prince George supports an unexpectedly high richness of phorid flies. Scuttle flies are especially abundant and diverse in greenbelt, edge, and residential sites. Temperature and land use significantly predict their abundance. Industrial areas are extremely hot and may offer fewer resources to flies. Conversely, many phorid mOTUs appear to specialize to greenbelt habitats. Because the life history of many species remains unknown, it can be hard to understand what factors drive these trends. 105 When looking at data from BOLD and GBIF, it is clear phorid richness has been largely unexplored in many parts of British Columbia. This project had added a suite of new datapoints to an underrepresented corner of the map, but much work remains to be done. More sampling is needed in the north and west of the province. New specimens should be vouchered, and taxonomists are needed to properly describe them and build a provincial species checklist. Although phorids prove to be a taxonomic challenge, DNA barcoding and new sampling efforts are helping us understand them better than ever before. APPENDIX I Tables for chapter two Table 5.1.1. Table of OTU guild affinities. Each guild is ranked as having no (0), weak (1), moderate (2) or strong (3) affinity to each feeding guild (Herb = herbivore, Nat. En. = Natural Enemy, Decomp. = Decomposer, Fung. = Fungivore, Hon. d.f. = Honeydew feeder). Lowest Classification Agyneta ordinaria Alydus sp. Amara idahoana Anoscopus sp. Anthomyiidae Anystidae Aphididae Aphrodes sp. Aphrophora sp. Arganthomyza duplex Atomaria sp. Basalys sp. Bathyphantes sp. Botanophila hucketti Braconidae Bradysia scabricornis Bradysia sp. Bradysia splendida Bradysia trivittata Herb. 0 3 3 3 NA 0 3 3 3 1 0 1 0 3 1 1 1 1 1 Nat. En. 3 0 2 0 NA 3 0 0 0 0 0 3 3 0 3 0 0 0 0 Decomp. 0 1 0 0 NA 0 0 0 0 3 0 0 0 0 0 3 3 3 3 Fung. 0 0 0 0 NA 0 0 0 0 0 3 0 0 0 0 2 2 2 2 Hon. d.f. 0 0 0 0 NA 0 0 0 0 0 0 0 0 1 1 0 0 0 0 106 Bryotropha similis Camnula pellucida Carabidae Carabus taedatus Cecidomyiidae Ceratagallia sp. Chaitophorus neglectus Chilopoda Chloropidae Chorthippus curtipennis Cicadellidae Cinara sp. Clubionidae Coccinella septempunctata Collembola Conioscinella sp. Cordyla sp. Corticarina cavicollis Corynoptera saccata Corythucha sp. Cryptophagus sp. Curculionidae Cybaeidae Cybaeus morosus Cytilus sericeus Delia sp. Devia prospera Dinotrema sp. Doratura stylata Entiminae Epipsocidae Eremocoris sp. Erythraeidae Exitianus sp. Forficula auricularia Formica sp. Gelis festinans Geocoris sp. Glocianus punctiger Gnaphosidae Habronattus ophrys Hahniidae Helina sp. Heliocobia rapax 3 3 NA 0 NA 3 3 0 NA 3 3 3 0 1 NA NA 0 0 1 3 0 NA 0 0 3 2 0 1 3 NA NA 3 0 3 2 1 1 1 3 0 0 0 1 1 0 0 3 3 NA 0 0 3 NA 0 0 0 3 3 NA NA 0 0 0 0 0 NA 3 3 0 0 3 3 0 NA NA 0 3 0 1 3 3 3 0 3 3 3 3 2 0 0 NA 0 NA 0 0 0 NA 0 0 0 0 0 NA NA 0 0 3 0 0 NA 0 0 0 3 1 0 0 NA NA 0 0 0 3 3 0 0 0 0 0 0 0 3 0 0 NA 0 NA 0 0 0 NA 0 0 0 0 1 NA NA 3 3 2 0 3 NA 0 0 0 0 1 0 0 NA NA 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 NA 0 0 0 NA 0 0 0 0 0 NA NA 0 0 0 0 0 NA 0 0 0 0 0 1 0 NA NA 0 0 0 0 3 1 0 0 0 0 0 0 1 107 Henicopidae Heterosilpha ramosa Hybotidae Hydrophoria sp. Incertella incerta Lamyctes emarginatus Larrinae Lasius sp. Leptocera sp. Leptothorax sp. Linyphiidae Liogluta nitens Lithobiidae Oedipodinae Lycosidae Manica sp. Megaselia Melanophthalma sp. Meteorus sp. Myrmica sp. Nearctaphis sensoriata Oscinella sp. Otiorhynchus sp. Oxypoda Parasitidae Pardosa moesta Pardosa tesquorum Phalangiidae Philodromidae Philonthus sp. Phrurolithidae Platygastridae Pollenia pediculata Polydesmidae Psammotettix confinis Pseudolycoriella sp. Psychidae Pterostichus adstrictus Pterostichus carbonarius Pterostichus melanarius Rhopalidae Sarcoptiformes Scaphinotus marginatus Schizolachnus sp. 0 0 0 NA 0 0 1 1 0 1 0 NA 0 3 0 1 NA 0 1 1 3 3 3 NA 0 0 0 1 0 0 0 1 1 0 3 NA 3 0 0 1 3 NA 0 3 3 3 3 NA 0 3 3 2 0 3 3 NA 3 0 3 3 NA 0 3 3 0 0 0 NA 3 3 3 3 3 3 3 3 3 0 0 NA NA 3 3 3 0 NA 3 0 0 3 0 NA 3 0 0 2 3 3 0 NA 0 0 0 3 NA 0 0 3 0 0 0 NA 0 0 0 1 0 0 0 0 0 3 0 NA NA 0 0 1 0 NA 0 0 0 0 0 NA 0 0 0 0 0 0 0 NA 0 0 0 0 NA 3 0 0 0 0 0 NA 0 0 0 0 0 0 0 0 0 0 0 NA NA 0 0 0 0 NA 0 0 0 0 0 NA 0 0 1 3 0 1 0 NA 0 0 0 1 NA 0 1 2 0 0 0 NA 0 0 0 0 0 0 0 1 0 0 0 NA 0 0 0 0 0 NA 0 0 108 Sciaridae Sciocoris microphthalmus Sciomyzidae Scotinella pugnata Silphidae Siphonella sp. Sitona hispidulus Slaterobius insignis Staphylinidae Stygnocoris sabulosus Stygnocoris sp. Synuchus impunctatus Tachyporus sp. Tenuiphantes zelatus Thaumatomyia sp. Thomisidae Trachelipus rathkii Trachyphloeus sp. Trapezonotus sp. Trichogramma sp. Tricimba sp. Trixoscelis sp. Tychius sp. Vespula pensylvanica Xysticus benefactor Xysticus montanensis Zelotes sp. 1 3 0 0 NA 1 3 3 NA 3 3 3 1 0 1 0 0 3 3 1 1 0 3 1 0 0 0 0 0 3 3 NA 3 0 0 NA 0 0 2 3 3 3 3 0 0 0 3 0 0 0 3 3 3 3 3 0 1 0 NA 0 0 0 NA 0 0 0 0 0 0 0 3 0 0 0 3 3 0 2 0 0 0 2 0 0 0 NA 0 0 0 NA 0 0 0 1 0 0 0 0 0 0 0 2 0 0 0 0 0 0 0 0 1 0 NA 0 0 0 NA 0 0 0 0 0 1 0 0 0 0 1 0 0 0 1 0 0 0 Table 5.1.2. OTU observations by site and land use type (Gb. = Greenbelt, Ind. = Industrial, Res. = Residential). Lowest Classification Agyneta ordinaria Alydus sp. Amara idahoana Anoscopus sp. Anthomyiidae Anystidae Aphididae Aphrodes sp. Aphrophora sp. Arganthomyza duplex Atomaria sp. Basalys sp. G1 0 0 0 1 0 0 1 0 0 0 0 1 Gb. G2 G3 1 0 0 0 0 0 0 0 2 0 1 1 1 0 0 0 0 0 0 0 2 0 0 1 G4 0 0 0 0 0 0 0 0 0 0 0 0 I1 0 0 0 2 7 0 22 0 0 0 1 0 Ind. I2 I3 0 0 1 0 1 0 0 0 0 0 0 2 6 2 0 1 0 0 1 0 5 0 0 0 I4 0 0 0 0 1 0 1 0 0 0 0 0 R1 0 0 0 1 0 1 3 0 1 0 0 0 Res. R2 R3 0 0 0 0 0 0 0 1 0 0 0 0 5 1 0 0 0 0 0 0 0 0 0 0 R4 0 0 0 0 1 1 3 0 0 2 1 0 109 Bathyphantes sp. Botanophila hucketti Braconidae Bradysia scabricornis Bradysia sp. Bradysia splendida Bradysia trivittata Bryotropha similis Camnula pellucida Carabidae Carabus taedatus Cecidomyiidae Ceratagallia sp. Chaitophorus neglectus Chilopoda Chloropidae Chorthippus curtipennis Cicadellidae Cinara sp. Clubionidae Coccinella septempunctata Collembola Conioscinella sp. Cordyla sp. Corticarina cavicollis Corynoptera saccata Corythucha sp. Cryptophagus sp Curculionidae Cybaeidae Cybaeus morosus Cytilus sericeus Delia sp. Devia prospera Dinotrema sp. Doratura stylata Entiminae Epipsocidae Eremocoris sp. Erythraeidae Exitianus sp. Forficula auricularia Formica sp. Gelis festinans Geocoris sp. Glocianus punctiger Gnaphosidae Habronattus ophrys 0 0 0 0 1 0 2 0 0 0 0 0 0 1 2 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 2 0 0 0 2 0 0 0 0 0 0 0 0 0 0 0 0 0 3 2 2 0 0 0 0 0 4 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 2 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 1 0 0 0 0 0 0 0 0 0 0 2 0 0 1 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 1 1 0 0 0 0 1 1 1 0 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 2 0 0 0 0 0 0 0 0 0 0 0 2 0 0 0 0 0 0 0 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 1 0 0 0 0 0 0 0 15 0 0 0 0 0 0 1 0 2 0 0 2 2 0 0 1 1 0 0 0 3 0 0 0 0 0 11 0 0 0 0 0 2 0 0 0 0 0 0 0 0 0 0 1 0 0 0 98 0 0 0 3 0 6 0 0 0 0 0 0 0 1 0 0 0 1 0 0 0 0 0 0 0 0 563 0 0 0 1 0 4 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 48 0 0 0 1 0 20 0 0 0 1 0 0 0 0 0 0 3 0 1 4 0 0 1 0 0 0 3 0 26 0 8 0 12 12 10 1 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 1 2 0 0 0 0 0 0 5 0 1 0 0 0 0 0 0 3 0 0 7 0 2 0 0 2 0 0 0 19 206 24 1 0 0 1 1 1 1 0 0 1 0 2 0 1 0 14 0 0 0 0 0 0 0 0 0 3 0 0 0 0 0 0 0 0 0 5 173 0 1 0 0 0 11 0 2 0 0 0 0 1 0 0 0 0 0 0 1 0 0 0 0 0 2 15 0 0 0 2 0 14 0 0 0 0 0 0 1 0 0 1 3 0 0 0 1 0 1 0 0 0 106 0 0 0 1 0 11 0 0 0 0 0 0 0 1 0 1 1 0 1 5 0 0 0 3 0 1 39 0 1 0 0 0 110 Hahniidae Helina sp. Heliocobia rapax Henicopidae Heterosilpha ramosa Hybotidae Hydrophoria sp. Incertella incerta Lamyctes emarginatus Larrinae Lasius sp. Leptocera sp. Leptothorax sp. Linyphiidae Liogluta nitens Lithobiidae Oedipodinae Lycosidae Manica sp. Megaselia sp Melanophthalma sp. Meteorus sp. Myrmica sp. Nearctaphis sensoriata Oscinella sp. Otiorhynchus sp. Oxypoda Parasitidae Pardosa moesta Pardosa tesquorum Phalangiidae Philodromidae Philonthus sp. Phrurolithidae Platygastridae Pollenia pediculata Polydesmidae Psammotettix confinis Pseudolycoriella sp. Psychidae Pterostichus adstrictus Pterostichus carbonarius Pterostichus melanarius Rhopalidae Sarcoptiformes 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 1 0 0 0 0 0 0 0 0 1 0 1 0 1 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 2 0 0 83 0 0 0 0 0 0 0 0 0 0 2 0 0 0 0 0 0 0 0 6 0 0 0 1 1 14 0 0 0 0 0 0 0 6 0 0 0 0 0 28 0 0 0 0 0 0 1 15 0 0 0 6 0 3 0 0 0 0 1 0 4 0 0 0 0 1 0 107 0 0 0 0 0 0 1341 11 0 1 0 1 0 3 0 0 4260 0 0 1 14 0 0 0 0 5 0 2 0 1 0 1 0 1 3 4 0 1 1 13 2 4 0 0 30 0 0 1 0 3 0 0 1 1 0 3 2 0 11 0 14 0 0 7 0 0 0 5 0 0 0 0 3 0 10 0 0 3 1 0 0 4 0 1 0 0 6 0 17 0 1 2 0 0 0 8 0 0 0 0 3 0 15 0 0 9 0 0 0 6 0 3 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 2 0 0 11 0 0 0 4 0 1 0 0 0 0 0 1 1 0 0 5 0 0 0 3 0 58 0 0 0 0 0 0 0 0 0 13 0 0 0 3 0 4 0 0 0 0 3 0 0 0 0 73 0 0 0 4 0 0 0 0 0 1 0 0 2 0 0 0 0 0 5 11 0 3 1 0 0 1 0 0 1 1 0 4 1 0 0 0 0 0 0 1 0 0 2 0 0 0 1 5 0 0 2 1 0 0 2 0 0 1 0 0 0 0 0 0 0 0 0 1 0 0 2 0 0 0 0 0 0 0 0 11 0 0 0 0 0 0 0 0 0 1 13 0 0 0 0 1 0 0 0 2 0 0 0 0 0 3 0 0 0 0 0 2 0 1 0 4 0 0 0 0 0 4 2 0 14 0 0 11 0 27 0 2 2 0 0 0 2 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 2 54 7 0 1 0 1 0 0 3 5 0 48 0 1 0 0 0 0 0 0 0 7 3 0 0 1 0 0 0 1 0 0 0 1 0 1 111 Scaphinotus marginatus Schizolachnus sp. Sciaridae Sciocoris microphthalmus Sciomyzidae Scotinella pugnata Silphidae Siphonella sp. Sitona hispidulus Slaterobius insignis Staphylinidae Stygnocoris sabulosus Stygnocoris sp. Synuchus impunctatus Tachyporus sp. Tenuiphantes zelatus Thaumatomyia sp. Thomisidae Trachelipus rathkii Trachyphloeus sp. Trapezonotus sp. Trichogramma sp. Tricimba sp. Trixoscelis sp. Tychius sp. Vespula pensylvanica Xysticus benefactor Xysticus montanensis Zelotes sp 34 0 0 0 0 0 0 0 0 0 0 0 0 1 0 1 0 0 0 1 0 5 0 0 0 1 0 0 0 1 1 0 0 0 0 0 0 0 0 0 0 0 2 0 0 0 0 0 1 0 0 1 0 0 1 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 4 0 0 0 1 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 10 0 1 0 0 1 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 1 1 4 0 0 0 0 0 0 0 0 0 0 0 1 1 0 0 0 0 0 0 0 0 0 3 0 0 0 0 0 0 0 0 8 0 0 0 0 0 1 0 0 0 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 3 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 2 6 0 1 48 0 0 0 0 0 0 4 0 4 1 0 4 1 0 26 0 0 1 0 0 0 0 0 22 0 0 0 0 0 3 1 0 0 0 0 0 6 0 2 0 0 3 1 0 0 0 1 0 0 0 0 0 0 10 1 0 1 0 1 0 0 0 0 1 0 0 1 0 9 0 0 2 0 0 6 0 0 0 1 0 0 0 5 1 0 0 1 0 0 0 0 0 0 0 0 1 0 0 2 0 0 1 0 0 52 0 0 0 112 Table 5.1.3. The association of OTUs with different land use types. I listed OTU’s habitat preference ranging from open areas (fields and grasslands) to closed (forests) as described in the literature. I tested OTU associations with different land use types using the rpb correlation index from the indicspecies package in R which conducts permutations (n=10000) to determine whether OTUs correlate more with groups of sites than randomly generated communities. In the land use column, one indicates that OTU is positively associated with a land use type while zero indicates it is not. The variable r represents the correlation for an OTU’s strongest habitat association (with a single land use type or a group of land use types), and p indicates the significance. The total number of positive associations are listed on the bottom of the table for each land use type. Species names are bolded when they have a 0.05 significance or lower. Lowest Classification Agyneta ordinaria Alydus sp. Amara idahoana Anoscopus sp. Anthomyiidae Anystidae Aphididae Aphrodes sp. Aphrophora sp. Arganthomyza duplex Atomaria sp. Basalys sp. Bathyphantes sp. Botanophila hucketti Braconidae Bradysia scabricornis Bradysia sp. Bradysia splendida Bradysia trivittata Bryotropha similis Camnula pellucida Carabidae Carabus taedatus Cecidomyiidae Ceratagallia sp. Chaitophorus neglectus Chilopoda Chloropidae Chorthippus curtipennis Cicadellidae Cinara sp. Habitat Preference Both Open Open Open Unknown Unknown Unknown Open Unknown Both Semi-Open Unknown Unknown Unknown Unknown Unknown Unknown Unknown Unknown Both Open Unknown Both Unknown Open Semi-Open Unknown Unknown Open Unknown Unknown Gb Ind Res 1 0 0 0 1 0 0 1 0 0 1 1 0 1 0 0 1 1 0 1 0 0 1 0 0 0 1 0 0 1 0 1 0 1 0 0 1 0 0 1 0 0 1 0 0 0 1 0 0 1 0 0 1 0 1 0 0 0 1 0 0 1 0 0 0 1 0 0 1 0 1 0 0 0 1 1 0 0 0 1 0 0 0 1 0 0 1 0 0 1 1 0 0 rpb 0.43 0.43 0.43 0.18 0.40 0.00 0.49 0.43 0.43 0.30 0.37 0.63 0.43 0.43 0.43 0.35 0.19 0.43 0.43 0.43 0.43 0.76 0.43 0.30 0.43 0.63 0.43 0.43 0.43 0.43 0.43 p 1.00 1.00 1.00 1.00 0.67 1.00 0.24 1.00 1.00 1.00 0.70 0.27 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 0.05 1.00 0.87 1.00 0.28 1.00 1.00 1.00 1.00 1.00 113 Clubionidae Coccinella septempunctata Collembola Conioscinella sp. Cordyla sp. Corticarina cavicollis Corynoptera saccata Corythucha sp. Cryptophagus sp Curculionidae Cybaeidae Cybaeus morosus Cytilus sericeus Delia sp. Devia prospera Dinotrema sp. Doratura stylata Entiminae Epipsocidae Eremocoris sp. Erythraeidae Exitianus sp. Forficula auricularia Formica sp. Gelis festinans Geocoris sp. Glocianus punctiger Gnaphosidae Habronattus ophrys Hahniidae Helina sp. Heliocobia rapax Henicopidae Heterosilpha ramosa Hybotidae Hydrophoria sp. Incertella incerta Lamyctes emarginatus Larrinae Lasius sp. Leptocera sp. Leptothorax sp. Linyphiidae Liogluta nitens Unknown 1 0 0 0.43 1.00 Both Unknown Unknown Unknown Open More Closed Unknown Closed Unknown Unknown Closed Open Open More Closed Unknown Open Open Closed Unknown Unknown Open More Open Unknown Open Open Open Unknown Open Both Unknown Both Unknown Semi-Open Unknown Both Open Both Unknown Both Both Unknown Unknown Both 0 0 0 0 0 0 1 1 0 1 1 0 0 1 1 0 0 1 1 0 0 0 1 0 0 0 0 0 1 0 0 0 0 1 0 0 0 0 0 0 0 1 0 1 1 1 0 1 1 0 0 0 0 0 0 1 0 0 1 0 0 1 1 1 0 0 1 1 1 1 1 0 1 1 1 0 0 1 1 1 1 0 1 1 0 0 0 1 0 1 0 0 0 0 1 1 0 1 1 0 0 1 1 0 0 0 0 1 0 0 0 0 0 0 0 1 0 0 1 1 0 0 0 0 1 0 0 0 1 0.43 0.48 0.43 0.43 0.43 0.43 0.43 0.43 0.25 0.32 0.43 0.68 0.52 0.59 0.19 0.48 0.43 0.43 0.35 0.43 0.43 0.66 0.33 0.43 0.47 0.43 0.39 0.43 0.43 0.30 0.43 0.43 0.43 0.32 0.41 0.43 0.43 0.43 0.97 0.82 0.43 0.21 0.43 1.00 0.29 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 0.05 0.28 0.27 1.00 0.28 1.00 1.00 0.67 0.49 1.00 0.05 0.69 1.00 0.06 1.00 0.63 1.00 1.00 1.00 1.00 1.00 1.00 1.00 0.71 1.00 1.00 1.00 0.01 0.05 0.10 0.84 1.00 114 Lithobiidae Oedipodinae Lycosidae Manica sp. Megaselia sp. Melanophthalma sp. Meteorus sp. Myrmica sp. Nearctaphis sensoriata Oscinella sp. Otiorhynchus sp. Oxypoda Parasitidae Pardosa moesta Pardosa tesquorum Phalangiidae Philodromidae Philonthus sp. Phrurolithidae Platygastridae Pollenia pediculata Polydesmidae Psammotettix confinis Pseudolycoriella sp. Psychidae Pterostichus adstrictus Pterostichus carbonarius Pterostichus melanarius Rhopalidae Sarcopteriformes Scaphinotus marginatus Schizolachnus sp. Sciaridae Sciocoris microphthalmus Sciomyzidae Scotinella pugnata Silphidae Siphonella sp. Sitona hispidulus Slaterobius insignis Staphylinidae Stygnocoris sabulosus Stygnocoris sp. Synuchus impunctatus Both Open Both Both Unknown Closed Unknown Both Semi-Open Open Both Closed Unknown More Open More Open Unknown Open Both Unknown Unknown Open Unknown Open Unknown Unknown Both More Open More Open Unknown Unknown Both Unknown Unknown 0 0 0 0 1 0 0 0 0 0 0 1 0 0 0 1 0 0 0 1 0 1 0 0 0 1 0 1 0 0 1 0 0 1 1 1 1 0 1 1 1 1 0 0 0 0 1 1 0 1 0 1 1 0 0 1 1 0 0 0 0 1 1 0 0 1 0 0 1 0 0 0 0 0 0 1 1 0 1 0 0 0 0 1 0 0 1 0 0 0 1 0 1 1 0 0 0 1 0 0.63 0.63 0.44 0.30 0.58 0.43 0.43 0.43 0.43 0.61 0.41 0.43 0.34 0.43 0.43 0.52 0.43 0.44 0.57 0.23 0.43 0.46 0.78 0.43 0.43 0.43 0.43 0.37 0.43 0.38 0.43 0.43 0.35 0.27 0.26 0.42 1.00 0.03 1.00 1.00 0.05 1.00 0.18 0.65 1.00 1.00 1.00 1.00 0.06 1.00 0.27 0.27 0.88 1.00 0.10 0.05 1.00 1.00 1.00 1.00 0.36 1.00 0.88 1.00 1.00 0.87 Both Closed Both Unknown More Open More Open More Open Unknown Semi-Open Unknown Semi-Open 0 1 0 0 1 0 0 1 1 0 1 1 0 1 0 0 0 1 0 0 0 0 0 0 0 1 0 1 0 1 1 1 0 0.43 0.43 0.53 0.43 0.43 0.43 0.43 0.50 0.32 0.63 0.43 1.00 1.00 0.27 1.00 1.00 1.00 1.00 0.43 1.00 0.27 1.00 115 Tachyporus sp. Tenuiphantes zelatus Thaumatomyia sp. Thomisidae Trachelipus rathkii Trachyphloeus sp. Trapezonotus sp. Trichogramma sp. Tricimba sp. Trixoscelis sp. Tychius sp. Vespula pensylvanica Xysticus benefactor Xysticus montanensis Zelotes sp Total Semi-Open Both Open Unknown More Open Open Open Unknown Both Open Open Both More Open Both Both 134 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 36 0 0 0 1 0 1 1 1 1 1 1 1 1 1 1 71 1 0 1 0 1 1 1 0 0 0 1 1 0 0 0 46 0.50 0.63 0.43 0.66 0.43 0.27 0.32 0.43 0.67 0.63 0.32 0.40 0.43 0.43 0.43 0.42 0.27 1.00 0.11 1.00 0.78 1.00 1.00 0.07 0.28 1.00 0.47 1.00 1.00 1.00 Table 5.1.4. Degree of host specialization in herbivorous taxa where 1 indicates feeding within a single genus of plants. 2 within a single family, and 3 within multiple families. Lowest Classification Alydus sp. Amara idahoana Anoscopus sp. Aphrodes sp. Aphrophora sp. Bryotropha similis Camnula pellucida Ceratagallia sp. Chaitophorus neglectus Chorthippus curtipennis Cinara sp. Cytilus sericeus Delia sp. Doratura stylata Eremocoris sp. Exitianus sp. Forficula auricularia Glocianus punctiger Nearctaphis sensoriata Oedipodinae Oscinella sp. Otiorhynchus sp. Host Specialization 3 3 2 3 3 3 3 3 1 2 1 3 3 2 3 3 3 1 3 3 2 3 116 Psammotettix confinis Psychidae Schizolachnus sp. Sitona hispidulus Slaterobius insignis Stygnocoris sabulosus Stygnocoris sp. Synuchus impunctatus Trachyphloeus sp. Trapezonotus sp. Tychius sp. 3 3 1 2 3 3 3 3 3 3 2 APPENDIX. II Tables for chapter three Table 5.2.1. BINs and species names from the BOLD identification engine that matched OTUs from Prince George. Individual counts (Ind Cnt) represent the number of specimens belonging to each mOTU, BIN, and species search result (left to right). Percent similarity (% sim.) indicates the mean % similarity between Prince George and BOLD sequences. Match count indicates the number of BOLD sequences that matched identifications (≥ 98.5%) (BINs and species). mOTUs highlighted in red matched more than one species. OTU-1 OTU-2 OTU-2 OTU-3 OTU-4 Ind Cnt 4 3 3 5 11 BOLD:ACV1885 BOLD:AEA0066 BOLD:AEA0067 BOLD:AAG3349 BOLD:AEU7203 Ind Cnt 1 2 2 2 3 OTU-4 11 BOLD:AEU7203 3 OTU-4 11 BOLD:AEU7203 3 OTU-5 OTU-6 OTU-7 OTU-8 4 2 4 17 BOLD:AAM9355 BOLD:AAP6420 BOLD:AAP6408 BOLD:ACB2325 2 1 3 14 OTU-9 58 BOLD:AAU8534 55 OTU-10 OTU-11 OTU-12 OTU-13 OTU-13 OTU-14 OTU-15 OTU-16 8 13 14 15 15 111 34 2 BOLD:ACX3812 BOLD:ABV3315 BOLD:AAU5599 BOLD:ADA4621 BOLD:ADA4621 BOLD:ACS7008 BOLD:ACX6055 BOLD:ACZ4030 1 5 13 14 14 94 27 1 mOTUs BIN Lowest taxonomic identification Aenigmatias Megaselia aequalis Megaselia giraudii Megaselia Megaselia brevicostalis Megaselia sp:BOLD:AAG3266 Megaselia sp. 40 SH2017 Megaselia devia Megaselia Megaselia pongsaiae Megaselia largifrontalis Megaselia pulicaria complex Megaselia Megaselia nigra Megaselia crassipes Megaselia giraudii Megaselia lucifrons Megaselia arcticae Megaselia brevicostalis Megaselia infraposita Ind Cnt 3 % sim. 99.8 99.4 99.8 100 98.7 Match Cnt 44 1 1 100 81-82 3 98.7 19 1 98.7 1 3 14 100 100.0 100 100 9 100 5 3-101 55 100 1-101 13 3 14 94 27 1 100.0 100 100 100 100 100 100 100 9 2 1-62 1 1-101 2-101 1-14 41 2 117 OTU-17 17 BOLD:ADA4916 14 Megaselia longicostalis 14 OTU-18 OTU-19 OTU-20 2 2 4 BOLD:ACB2539 BOLD:AAU6605 BOLD:ACC5889 1 1 2 1 1 OTU-21 7 BOLD:AAL9073 6 OTU-22 OTU-23 10 2 BOLD:ACF3749 BOLD:AAN8687 2 1 Megaselia hendersoni Megaselia meconicera Megaselia Megaselia pulicaria complex Megaselia Megaselia fujiokai OTU-24 7 BOLD:AAP6413 2 Megaselia OTU-25 OTU-26 OTU-27 OTU-28 OTU-29 OTU-30 OTU-31 3 1 1 18 1 3 3 BOLD:AAL9075 BOLD:AEV9051 BOLD:ACX1916 BOLD:AAG3274 BOLD:AAP8725 BOLD:ABU5538 BOLD:AAG3259 3 1 1 16 1 2 2 Megaselia lucifrons Megaselia Megaselia Megaselia hardingorum Megaselia Megaselia Megaselia longipennis OTU-32 9 BOLD:AAG3315 3 Megaselia OTU-33 1 BOLD:AAG3310 1 Megaselia OTU-34 8 BOLD:AEV9050 4 Megaselia OTU-35 OTU-36 OTU-37 OTU-38 OTU-39 OTU-40 OTU-41 OTU-42 1 2 5 1 12 9 2 1 BOLD:AAG3286 BOLD:ACD4564 BOLD:AAU6624 BOLD:ACB0962 BOLD:ABA6993 BOLD:AAG3266 1 2 3 1 12 3 BOLD:AAG3343 1 OTU-43 2 BOLD:AAG3338 1 Megaselia lombardorum Megaselia Megaselia spinicincta Megaselia citrinella Megaselia pleuralis Megaselia Megaselia Megaselia Lecanocerus compressiceps OTU-44 3 BOLD:AAG3238 2 Phora OTU-45 OTU-46 OTU-47 OTU-48 OTU-49 OTU-50 OTU-51 OTU-52 OTU-53 OTU-54 OTU-55 OTU-56 OTU-57 OTU-58 OTU-59 1 3 2 1 3 2 1 16 2 2 2 3 5 1 2 BOLD:ABY0932 BOLD:ACX6289 BOLD:AAU5646 BOLD:AAU5624 BOLD:AAG3236 BOLD:AAP4685 BOLD:ACS2187 BOLD:AAU6533 BOLD:AAU5652 BOLD:AAN8696 BOLD:ADP3479 BOLD:ABX8608 BOLD:ADF8082 BOLD:AAL9070 BOLD:AAG3235 1 3 1 1 3 1 1 13 1 2 2 2 5 1 1 Megaselia Metopina galeata Puliciphora Pseudoaceton Diplonevra nitidula Megaselia Megaselia Megaselia coaetanea Megaselia Megaselia Megaselia rufipes Megaselia thomseni Megaselia longiseta Megaselia 6 1 3 16 2 1 3 1 12 1 3 3 13 2 2 1 100 11101 98.6 100.0 1 48 100 8-101 100 100 99.9'100 100 99.6 99.5 100 99.7 100 99.5 99.199.8 99.2 99.8'100 99.6 99.4 100 100 100 100 100 25 100 101 100 3 1-101 43 101 1 80-96 100 100 100 17 1 101 101 101 98.9 5 100 12 99.8'100 99.7 100 99.7 99.5 100 99.7 100.0 100 99.7 99.8 100 98.6 100 99.8 100 78 26 28 100 101 67 17 5 13 39 19 2-4 2 3 118 OTU-60 3 BOLD:AAG3304 3 Megaselia nigriceps OTU-61 OTU-62 OTU-63 OTU-64 OTU-65 OTU-66 OTU-67 OTU-68 OTU-69 OTU-70 OTU-71 OTU-72 OTU-73 OTU-74 OTU-75 1 2 3 1 11 1 1 2 3 1 1 3 2 1 2 BOLD:ACA2662 BOLD:AAM9373 BOLD:AAZ6701 BOLD:ACL8521 BOLD:AAG3356 BOLD:ABX8135 BOLD:AAG3235 BOLD:AAG3355 BOLD:ACB0987 1 BOLD:ACP5869 BOLD:ACB0673 1 2 Megaselia Megaselia Megaselia joanneae Megaselia Megaselia Megaselia Megaselia Megaselia Megaselia spinigera Megaselia Megaselia Megaselia Metopina-group Megaselia gregaria Megaselia stoakesi OTU-76 2 BOLD:AAG3311 2 Megaselia tecticauda OTU-77 OTU-78 OTU-79 OTU-80 OTU-81 OTU-82 OTU-83 OTU-84 OTU-85 OTU-86 OTU-87 OTU-88 OTU-89 OTU-90 OTU-91 OTU-92 OTU-93 OTU-94 OTU-95 OTU-96 4 1 2 2 1 1 1 1 1 1 1 1 2 1 1 1 1 19 4 3 OTU-96 3 1 2 1 1 1 3 BOLD:ACN5785 1 BOLD:ABX8427 BOLD:AAG3302 1 1 BOLD:AAG3351 1 BOLD:AAP8118 BOLD:AAM9346 1 1 BOLD:ACB6729 BOLD:ACB6729 BOLD:AAU6529 11 2 2 3 BOLD:AAU6529 2 OTU-97 3 BOLD:AAU6529 3 OTU-97 3 BOLD:AAU6529 3 OTU-98 OTU-99 1 1 BOLD:ACC7814 BOLD:ACC7711 1 1 Megaselia Megaselia Megaselia devia Megaselia Megaselia Megaselia Megaselia atrox Megaselia losangelensis Phora Megaselia Megaselia lutea Megaselia Gymnophora subarcuata Megaselia fungivora Megaselia Megaselia Megaselia Megaselia cinereifrons Megaselia Megaselia baileyae Megaselia pulicaria complex Megaselia baileyae Megaselia pulicaria complex Megaselia Megaselia 3 100 1'100 12 23 8-56 4 100 56 100 44 53 3 99.9 99.5 100 100.0 100.0 99.7 99.9 99.8 100 1 2 99.3 100 2 100 1 99.4 2 1 1 99.9 99.4 101 32 1 100 36 1 1 100 100 11 23 11 2 100 99.7 100 2 68 2 2 100 1 3 98.7 33-34 2 98.6 1 99.6 99.9 57 79 3 101 71 16101 119 Table 5.2.2. Number of phorid flies found in each land use type (Greenbelt = green, Edge = orange, Residential = blue, Industrial = gray), by year (2015 = 15, 2022 = 22) and the results of habitat associations found by the indicspecies package function multipatt. The rpb index represents the correlation between mOTUs and groups of sites. It displays the land use type or groups of land use types that had the most significant association (p) for each mOTU. mOTUs with significant associations were bolded. The zero and one values in the right section of the table indicate no or positive associations. mOTUs that occur in all land use types are highlighted in yellow, while those that occur in only one land use type are highlighted in purple (Note: some are singletons). mOTUs OTU-1 OTU-2 OTU-3 OTU-4 OTU-5 OTU-6 OTU-7 OTU-8 OTU-9 OTU-10 OTU-11 OTU-12 OTU-13 OTU-14 OTU-15 OTU-16 OTU-17 OTU-18 OTU-19 OTU-20 OTU-21 OTU-22 OTU-23 OTU-24 OTU-25 OTU-26 OTU-27 OTU-28 OTU-29 OTU-30 OTU-31 OTU-32 15 0 1 2 1 1 1 1 9 35 7 2 14 3 81 0 1 5 1 1 2 2 5 0 3 1 0 1 2 1 2 2 4 22 22 15 22 15 22 Gb Edg Res Ind rpb p 0 0 2 0 1 1 0 0 1 1 0.27 0.55 0 0 0 0 0 2 1 0 0 1 0.22 0.79 3 0 0 0 0 0 1 0 0 0 0.47 0.07 1 2 0 4 0 3 0 1 0 0 0.33 0.24 0 0 0 0 1 2 0 0 0 1 0.32 0.26 1 0 0 0 0 0 1 0 0 0 0.35 0.39 0 0 0 3 0 0 0 0 1 0 0.34 0.11 8 0 0 0 0 0 1 0 0 0 0.59 0.01 8 1 0 10 2 2 1 0 0 0 0.27 0.21 0 0 0 1 0 0 1 0 0 0 0.30 0.25 0 0 0 7 1 3 0 0 1 0 0.25 0.46 0 0 0 0 0 0 1 0 0 0 0.24 1.00 8 1 0 3 0 0 1 0 0 0 0.38 0.12 9 2 0 14 2 3 1 0 0 0 0.24 0.62 0 0 7 9 3 15 0 0 1 1 0.48 0.03 0 1 0 0 0 0 0 1 0 0 0.47 0.11 6 1 0 5 0 0 1 1 1 0 0.39 0.08 0 0 0 1 0 0 1 0 1 0 0.19 0.78 0 1 0 0 0 0 0 1 0 0 0.47 0.11 0 1 0 0 0 1 0 1 0 0 0.24 0.60 4 1 0 0 0 0 1 1 0 0 0.47 0.02 1 0 1 0 1 2 1 0 0 0 0.25 0.49 0 0 1 0 0 1 0 0 1 1 0.19 0.78 3 0 0 1 0 0 1 0 0 0 0.42 0.07 0 0 0 1 0 1 1 0 1 1 0.07 1.00 0 0 0 0 1 0 0 0 0 1 0.24 1.00 0 0 0 0 0 0 1 0 0 0 0.24 1.00 2 1 0 12 0 1 0 0 1 0 0.35 0.16 0 0 0 0 0 0 1 0 0 0 0.24 1.00 1 0 0 0 0 0 1 0 0 0 0.43 0.09 0 0 0 1 0 0 1 0 1 0 0.18 0.79 2 2 0 1 0 0 1 1 0 0 0.40 0.09 120 OTU-33 OTU-34 OTU-35 OTU-36 OTU-37 OTU-38 OTU-39 OTU-40 OTU-41 OTU-42 OTU-43 OTU-44 OTU-45 OTU-46 OTU-47 OTU-48 OTU-49 OTU-50 OTU-51 OTU-52 OTU-53 OTU-54 OTU-55 OTU-56 OTU-57 OTU-58 OTU-59 OTU-60 OTU-61 OTU-62 OTU-63 OTU-64 OTU-65 OTU-66 OTU-67 OTU-68 OTU-69 OTU-70 OTU-71 OTU-72 OTU-73 OTU-74 OTU-75 OTU-76 1 1 1 2 1 1 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 5 0 0 0 0 0 7 0 1 1 2 1 0 0 0 0 1 1 0 1 0 0 0 0 0 2 0 0 0 0 0 7 0 0 1 0 1 1 0 2 0 1 0 0 2 0 0 0 0 0 0 0 0 1 0 0 0 0 1 1 0 0 3 0 0 0 0 1 0 0 0 0 0 1 1 3 0 0 1 0 0 0 0 0 0 1 1 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 4 0 12 0 1 0 0 1 0 0 1 0 2 1 0 7 0 2 2 2 0 1 0 2 1 2 2 0 1 1 0 0 2 0 0 0 0 1 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 3 1 0 0 0 0 6 1 0 0 1 4 0 0 1 0 0 0 0 0 0 1 0 1 0 0 3 0 0 0 0 1 1 1 1 0 1 0 1 0 1 0 1 1 0 0 0 0 1 1 0 1 0 0 0 0 0 1 0 0 0 0 0 1 0 0 0 0 1 1 0 1 0 0 0 0 1 0 0 0 0 0 0 0 0 1 0 0 0 0 1 1 0 0 1 0 0 0 0 1 0 0 0 0 0 1 1 1 0 0 1 0 0 0 0 0 0 1 1 0 0 0 0 1 0 1 0 1 0 0 1 0 0 1 0 1 1 0 1 0 1 1 1 0 1 0 1 1 1 1 0 0 1 0 0 1 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 0 0 0 0 1 1 0 0 0 1 0 0 0 0 0 0 0 0 0 1 0 0 0 0 1 0 0 0 0 0.24 0.45 0.24 0.24 0.31 0.24 0.43 0.39 0.37 0.24 0.47 0.23 0.24 0.43 0.19 0.70 0.31 0.19 0.24 0.40 0.18 0.37 0.26 0.24 0.33 0.26 0.24 0.24 0.26 0.26 0.40 0.70 0.41 0.26 0.24 0.47 0.24 0.24 0.24 0.24 0.24 0.26 0.47 0.47 1.00 0.07 1.00 1.00 0.26 1.00 0.07 0.11 0.18 1.00 0.10 0.59 1.00 0.08 0.79 0.06 0.35 0.79 1.00 0.07 1.00 0.19 0.35 0.40 0.16 0.36 1.00 0.40 0.35 0.36 0.10 0.05 0.08 0.36 1.00 0.11 0.41 1.00 1.00 1.00 1.00 0.36 0.10 0.10 121 OTU-77 OTU-78 OTU-79 OTU-80 OTU-81 OTU-82 OTU-83 OTU-84 OTU-85 OTU-86 OTU-87 OTU-88 OTU-89 OTU-90 OTU-91 OTU-92 OTU-93 OTU-94 OTU-95 OTU-96 OTU-97 OTU-98 OTU-99 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 4 0 2 0 0 0 1 0 0 2 1 0 0 0 1 1 1 1 2 1 1 1 1 4 3 0 2 1 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 2 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 2 0 2 0 0 0 1 0 0 0 0 0 0 0 0 0 0 8 1 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 1 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 0 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 1 0 0 0 0 0 1 0 0 0 1 0 0 0 0 0 0 0 0 0 0 1 0 0 1 0 0 0 0 0 0 0 1 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.30 0.70 0.26 0.35 0.24 0.24 0.26 0.24 0.24 0.24 0.24 0.24 0.35 0.24 0.24 0.24 0.24 0.31 0.25 0.29 0.23 0.24 0.24 0.45 0.05 0.35 0.38 1.00 1.00 0.35 1.00 1.00 1.00 1.00 1.00 0.38 1.00 1.00 1.00 1.00 0.28 0.53 0.39 0.59 1.00 1.00 122