C H A R T I N G B I O D I V E R S I T Y – S C U T T L E F L I E S A N D O T H E R P O O R L Y K N O W N I N S E C T S I N S W E D E N

Sibylle Häggqvist

Charting biodiversity Scuttle flies and other poorly known insects in Sweden

Sibylle Häggqvist

© Sibylle Häggqvist, Stockholm University 2016 Front cover illustration of Megaselia sp. by Martin Holmer Divider illustrations by Lothar Noack ISBN print 978-91-7649-501-8 ISBN PDF 978-91-7649-502-5 Printed in Sweden by Holmbergs, Malmö 2016 Distributor: Department of Zoology, Stockholm University

To my children

Abstract

Biodiversity has fascinated people in all times. We as a species are part of the global ecosystem and our survival depends on many other species around us. Charting biodiversity, however, has proven more difficult than one could imagine. When Linnaeus and others started describing which species there are on Earth, people in general, and taxonomists in particular, could not imagine that we would not be finished 300 years later and even less that we would not be able to tell exactly how much there is left to describe. In this PhD thesis, I deal with relatively large organisms (compared to the average size of organisms left to be described), namely insects in general and scuttle flies in particular, within a limited and well-studied geographic region (Sweden). Nevertheless, the results show that we are far from completing the inventory for even this limited portion of global biodiversity. Since it was in Sweden that Linnaeus started his work and where he did most of it, the Swedish flora and fauna belong to the best known in the world. In spite of this, we show in paper I of this thesis that it is likely that a considerable portion of the Swedish insect fauna remains to be discovered. The white spots on the biodiversity map primarily concern small Diptera and Hymenoptera species that are decomposers or parasitoids. The scuttle flies (Diptera: Phoridae) are one of the insect families that contain more undescribed species than species known to science, both in Sweden and elsewhere. The situation is complicated by the fact that a large portion of the known scuttle-fly diversity is placed within a single genus, Megaselia, one of the most species-rich genera in the animal kingdom; almost 40 % of described phorids belong to this genus. Our lack of understanding of phylogenetic relationships within Megaselia has made taxonomic progress on the genus very difficult. In paper II, we define a new natural clade within Megaselia, the lucifrons group, based on both molecular and morphological evidence. We show that the group comprises at least three species in Sweden, one of which is described by us as new to science. In paper III, we present the first comprehensive molecular phylogeny of Megaselia. We identify 22 well-supported natural clades within the genus, comprising at least two species each. These clades are recognized as species groups; most of them have not been described previously. We identify eight additional, isolated single-species lineages, which may turn out to represent multi-species clades when molecular data become available for more species. The paper includes the description of 45 species new to science, all from

Sweden. In paper IV, we revise the most basal of the 22 species groups, the spinigera group, and describe one additional new species to science from Sweden in this group. The thesis provides the first reasonably complete phylogenetic framework for Megaselia and its closest relatives, greatly facilitating further research into scuttle-fly diversity in Sweden and elsewhere.

The thesis is based on the following articles, which are referred to in the text by their Roman numerals:

I. Ronquist F, Häggqvist S, Forshage M, Karlsson D, Jaschhof M, Britton T, Kjærandsen J, Hovmöller R, Bergsten J, Gärdenfors U. Size and composition of the Swedish insect fauna. Manuscript II. Häggqvist S, Ulefors SO, Ronquist F (2015) A new species group in Megaselia, the lucifrons group, with description of a new species (Diptera, Phoridae). ZooKeys 512: 89–108. doi: 10.3897/zookeys.512.9494 III. Häggqvist S, Ulefors SO, Ronquist F. Phylogeny and species-group classification of the mega-diverse genus Megaselia (Diptera, Phoridae). Submitted to Systematic Entomology. IV. Häggqvist S, Ulefors SO, Ronquist F. The spinigera group of Megaselia (Diptera, Phoridae): molecular phylogeny, revision of the Swedish fauna, and description of a new species. Manuscript. Paper II is open access and is herein reproduced under the permission of the creative commons by attribution license (CC-BY). * The nomenclatural acts in this thesis are not issued for permanent scientific records or for purposes of Zoological Nomenclature and are not regarded as published within the meaning of the International Code of Zoological nomenclature (ICZN, ed. 4, article 8.2).

Candidate contributions to thesis articles*

I II III IV Conceived the study

- significant substantial substantial

Designed the study

- substantial substantial substantial

Collected the data

significant significant significant substantial

Analyzed the data

significant substantial substantial substantial

Manuscript preparation

minor substantial substantial substantial

* Contribution Explanation

Minor: contributed in some way, but contribution was limited. Significant: provided a significant contribution to the work. Substantial: took the lead role and performed the majority of the work.

Contents

Introduction .................................................................................................. 13

Material and methods .................................................................................. 22Material ................................................................................................................... 22Species richness and faunal composition ............................................................... 23Morphological methods ........................................................................................... 25Molecular methods .................................................................................................. 26Data analyses ......................................................................................................... 26

Results ......................................................................................................... 27Paper I ..................................................................................................................... 27Paper II .................................................................................................................... 28Paper III ................................................................................................................... 29Paper IV .................................................................................................................. 31

Discussion .................................................................................................... 33

Conclusion ................................................................................................... 38

Svensk Sammanfattning .............................................................................. 40

Acknowledgements ...................................................................................... 43

References ................................................................................................... 46

13

Introduction

“We shall not cease from exploration, and the end of all our exploring will be to arrive where we started and know the place for the first time.” – T. S. Eliot Knowing a place for the first time in terms of biodiversity or in other words describing the species in the world around us has been an ongoing task since the days of Linnaeus and is one of the fundamental quests in biology (Agnarsson and Kuntner 2007). The motivation for doing so may be different for different people. Some may argue that it is interesting to know what other life surrounds us, simply another step in the human striving for knowledge. Some may argue that it is vital to know what other species there are for reasons like competition for resources with other species than our own or in order to know which species can be used as resources. Others again may say that we have an obligation to know what is around us - if only to document what organisms are disappearing because of the ever-rising number of human beings on this planet and the destruction of ecosystems that comes with it (Dunn, 2005). Regardless of the justification, describing or even determining the number of species in a given habitat is not an easy task. When can we ever be sure that there are no more species that we could find if we continued to look a little longer? Physically finding each and every species in a given area is typically not a task that is feasible with a realistic investment of resources. Instead we need to aim at finding as many species as possible in the most effective way (Azevedo et al. 2013). To help us when knowledge is incomplete, researchers have developed statistical ways of estimating the number of species in a given habitat from a subsample of species. Different estimators have different strengths and weaknesses and can be used for different kinds of data (Hortal et al. 2006). Some of them are used in this thesis. Eventually, however, we do need to try to find and describe all species in order to fully appreciate the biological diversity of the planet. At the same time as researchers are working towards describing all species on Earth, we are facing what some call the sixth mass extinction. According to Ceballos et al. (2015), the rate of extinction that we can measure today

14

greatly exceeds the background rates of extinction that occurred between the other five mass extinction events in the history of life on Earth. It is suggested and widely accepted that this sixth mass extinction event is due to human activity, and a new geological age – the Anthropocene – starting with the industrialization has been recognized in order to stress the importance of human societies as a geophysical force (Steffen et al. 2007). The disappearance of megafauna from all continents except Africa can serve as an example: it has been suggested that these events depend to a large extent on human activity. For instance, data from a burnt eggshell of Genyornis newtoni, a now extinct, flightless 200 kg bird in Australia, gives direct evidence that the eggs of this species were preyed on by humans (Miller et al. 2016). In order to stop human-related mass extinction, conservation measures may still be effective but time is running out (Ceballos et al. 2015). One aspect, which is often neglected due to the difficulty to accurately evaluate the data we have, is the impact of human activity on invertebrates. While it is relatively easy to collect sufficient data on large well-known vertebrates, it is harder to collect data on the conservation status of relatively unknown or even undescribed invertebrates. Recently, Régnier et al. (2015) suggested that the global level of recently extinct species may be as high as 7 % when invertebrate data are factored in appropriately, as opposed to the 1.3 % that are suggested from looking at the IUCN Red List. Thus, the authors emphasize the importance of trying to evaluate the conservation status of invertebrates even before the complete fauna is described. Describing the invertebrate fauna is a daunting task: even if we could speed up the process of describing new species, we might well have another two and a half centuries of taxonomic work in front of us before the inventory of the invertebrate fauna of the planet is complete. In terrestrial habitats, insects constitute a major part of the unknown invertebrate fauna. A generally accepted view is that the approximately 1 million species of insects that are currently known to science comprise only about 10-20 % of the actual global insect fauna (Scheffers et al. 2012). Not surprisingly, some regions are better studied than others. In general, one can argue that Europe is the best-studied region globally partly due to its long tradition of taxonomic research and partly because of its temperate climate. Temperate regions are well known to have lower species richness than the tropics (Gillman et al. 2015), and temperate faunas should therefore be more completely known all else being equal. Even though it is not clear whether the latitudinal gradient of species richness actually holds for all taxonomic groups of insects, it is decidedly the general trend (Hillebrand 2004). However, as some researchers argue (e.g. Fontaine et al. 2012), even though Europe is a well-studied region, new species are still discovered at unprecedented high rates.

15

If it is difficult to estimate the true number of insect species for a given region or globally, it may be even harder to give an exact picture of the composition of the insect fauna. It appears highly likely that the proportion of species of a taxonomic group that is described is dependent, at least partly, on the ecological niche that the group fills. We probably already know a relatively large part of the phytophagous insects because we have been interested in the potential pests on our crops. Pollinators and predatory insects are probably also among the better-known ones because it is easy to see direct impacts on economically important crops and/or their enemies. Saprophytes and parasitic insects are less well known, partly because we have been considering them less relevant for agriculture but also because they are physically harder to find. Saprophytes and parasites often live inside the substrate they are feeding on and are thus more hidden from human observers. Our biased knowledge of insect groups may cause problems in many contexts. For instance, Shaw and Hochberg (2001) argue that the poorly known parasitoid Hymenoptera should be considered for conservation to a much larger extent than they are currently. In Britain, having one of the best known insect faunas of the world, parasitoid Hymenoptera comprise about 25% of the total insect diversity, but no conservation strategies for those species exist today (Shaw and Hochberg 2001). Traditionally, insect collectors have focused their interest on Lepidoptera (butterflies and moths), especially the butterflies, and secondly on Coleoptera (beetles), arguably the two orders containing the most strikingly beautiful insects. Thus, of the mega-diverse orders of insects (Coleoptera, Diptera, Hymenoptera, Lepidoptera), Hymenoptera (wasps, ants and bees) and Diptera (flies and midges) are the least known. Particularly species-rich and poorly known families of these orders include but are not restricted to scuttle flies, fungus gnats, gall midges (in Sweden at least until recently) and most families of parasitic wasps. These tend to be relatively small insects that often also have an inconspicuous lifestyle. So far not so many inventories of entire insect faunas have been attempted. The Swedish Malaise Trap Project (SMTP) may be the first inventory that is aiming at complete coverage of the insect diversity of a country. The SMTP forms part of the Swedish Taxonomy Initiative (STI), a project launched in 2002 with the aim of completing the inventory of the country’s flora and fauna in 20 years. During 2003-2006, SMTP collected approximately 80 million insect specimens with Malaise traps at some 50 sites spread across the country, and the material is now being sorted into fractions in order to make it available to taxonomic study by specialists around the world (Karlsson et al. 2005). An insect diversity sampling project that was inspired by the SMTP is the Estonian Malaise Trap Project (EMTP), which was in its

16

collecting phase during 2008-2011 and whose first results are now being published (Soon et al. 2011, Tomasson et al. 2014). René Malaise at the Swedish Museum of Natural History invented the Malaise trap (Figure 1) in 1934 (Vårdal and Taeger 2011). Over time, it proved to be a hugely successful trap for catching a variety of flying insects. When René Malaise, during an expedition to Burma, lay in his tent watching insects escaping through a tiny hole in the roof and came up with the basic design for the trap, did he imagine that his invention would still be used 80 years later? The popularity of his trap stems from the fact that it is easy to assemble and operate. It is also timesaving, as it does not require the collector herself to be present at the collecting site. The trap is working for her on its own; the only thing she has to do is empty it every now and then and, of course, to sort through the catch.

Figure 1: Malaise trap south of Stockholm.

Malaise traps are effective for a surprisingly large range of flying insects (Brown 2005). Typically, no baits are used; instead, Malaise traps work as flight interception traps, in the same way as windowpane traps. However, windowpane traps allow more insects to escape than Malaise traps do; therefore, Malaise traps generally catch a larger proportion of the fauna of

17

flying insects. It has been shown that windowpane traps are more efficient for collecting beetles and cockroaches than Malaise traps, whereas Malaise traps are more efficient for collecting Diptera, Hymenoptera and Hemiptera (Lamarre et al. 2012). Other traps like yellow pan traps or light traps make use of the fact that certain insects are attracted to certain colors or light. Insects can also be attracted to traps using different kinds of baits (e.g. food or pheromones) that emit olfactory signals that are attractive to certain species. For general collection of poorly known insect groups in the orders Hymenoptera and Diptera, however, Malaise traps are unsurpassed. In Sweden, poorly known insect groups are almost exclusively found in the Hymenoptera and Diptera (paper I); therefore, Malaise traps were the natural choice for the Swedish insect fauna inventory (Karlsson et al. 2005). In temperate regions such as Sweden, the traps need to be emptied every second week; even during the summer, more frequent emptying is usually not necessary. This means that the traps are easy to operate, but the work after the collection of the insects is still quite challenging and requires many man-hours of work. It involves sorting the collected specimens into suitable fractions that can be sent to specialists for species identification. The returned information and material then needs to be processed, and the data made available to the scientific community. Usually, researchers only pick out their group of interest, possibly one or two other groups for colleagues, but leave the rest of the Malaise trap catch untouched. Here, the SMTP is unique in its ambition to identify the entire collected material, regardless of taxonomic group. In paper I, we use data from the SMTP as a sample of the Swedish insect fauna, and use statistical methods for estimating the total diversity of the fauna. We compare these results with estimates by taxonomic experts of the proportion of the Swedish insect fauna that is till unknown. Our results indicate that as much as 20 % of the Swedish insect fauna may remain to be discovered. As expected, many of the new species are expected in small Diptera and Hymenoptera that are decomposers or parasitoids. Despite the lack of knowledge about them, many of these groups are very probably important within their ecosystems, and some species may even be key players. The scuttle flies (Diptera: Phoridae), the main focus of papers II-IV of this thesis, are a prime example of this. They belong to the most common insect groups in Malaise traps, despite the fact that other trap types are known to be even more effective in collecting these small flies. The variation of biology within the Phoridae and their abundance in most habitats suggest that they are important in the functioning of our ecosystems.

18

Today, more than 4,200 species of phorids are described (www.phorid.net/pcat/), but this may represent as little as 10 % of their actual diversity (Bonet et al. 2011). Phoridae occupy an astonishing variety of ecological niches. While many species are parasites or parasitoids of hymenopterans or spiders, and some are phytophagous, the majority of species appears to be saprophytic (Disney, 1994). Brown et al. (2015) suggest that many species have extremely specialized life histories, explaining how large numbers of closely related species can co-occur in the same habitats. However, for most species, including virtually the entire genus Megaselia, the biology is only poorly known or not known at all. There are several possible reasons as to why phorids are understudied. Typically, many phorid species co-occur in the same habitat, and these species may be difficult to differentiate, especially for the beginner. Good identification literature is scarce or, for many faunas, not available at all, making the identification quite challenging. Very few people know the group well; thus, it is difficult to get expert help with identification. Although the cynic could see an advantage in that at least not too many researchers have “messed up” the taxonomy of phorids, this is more than outweighed by the scarce literature and lack of experts on the group. A huge obstacle to progress in studying scuttle-fly diversity is the lack of knowledge about phylogenetic relationships. The single genus Megaselia comprises more than 1,600 described species, about 40 % of all known phorids. This makes it one of the largest genera across all of life1. In the 1 In the plants, Astragalus L. (Fabaceae), with 3,200 described species, is the largest and Bulbophyllum with 2,000 species the second largest genus (Frodin 2004). Within the mushroom-forming fungi, Cortinarius is the largest with 2,000 estimated species worldwide (Frøslev et al. 2007). In the animal kingdom, the genus Agrilus (Coleoptera, Buprestidae) with over 3,000 described species is considered to be the most species rich (Jendek and Poláková 2014). However, considering that 85% of the species of flowering plants are known (Joppa et al. 2011) and that beetles are generally more well-known than flies, while it is estimated that as little as 10-15 % of the Phoridae may be described so far (Bonet et al. 2011), it is quite possible that Megaselia is the largest genus across all of life. It is certainly one of the biggest. However, these comparisons are of course subject to change depending on the circumscription of genera in different organism groups. We are far from an objective criterion for circumscribing genera, so current boundaries are partly based on ease of recognizing the clades, whether they are known to be natural, and on practical considerations. From a practical perspective, it would not be surprising to see most of these megadiverse genera, including Megaselia, eventually split into a number of smaller genera.

19

Swedish fauna, the dominance of Megaselia is even more pronounced than globally, about 90% of the Swedish phorid species belong to Megaselia (Figure 2). Many new phorid species are described simply as Megaselia, without detailed references to their affinities to subgroups within the genus. Indeed, Megaselia subgroups remain poorly defined, if they have been discussed in the literature at all, and Megaselia is in dire need of a thorough taxonomic revision. However, taking on the entire group, even in just a limited geographic area, is not an easy task. Megaselia was earlier divided into two subgenera, Megaselia and Aphiochaeta (Schmitz 1926, 1927) based on whether or not hairs are present on the anepisternum. However, Disney (Weinmann and Disney 1997) abolished the subgenus Aphiochaeta formally based on a pair of supposedly closely related species where one species has a hairy anepisternum while it is bare in the other species. The monophyly of the subgenera has not been investigated further since then. Schmitz made further attempts to divide Megaselia into smaller divisions (“Abteilungen” and “Reihen”) using morphological characters on the thorax and in the wings (Schmitz 1956). However, his revision of the genus Megaselia was never completed and the groups he identified remain untested by other researchers.

Figure 2: Habitus of a typical Megaselia male illustrated by Martin Holmer.

20

Molecular methods have revolutionized insect taxonomy. Not only can new species be discovered and morphological findings tested, but the relationships among species and higher groups can also be studied. Another advantage of using molecular markers is that it opens up the possibility to reliably find both sexes of one species, even when species-specific characters are only known from one sex (Cook and Mostovski 2002). This can be important in groups like the Phoridae, where much of the taxonomic work is focused on male genitalia, and species descriptions and keys often do not include females at all. Despite their great potential, molecular methods had not been applied widely to phorids before this thesis (for exceptions, see Brown and Smith 2010, Smith and Brown 2010, Hash et al. 2011, Hash and Brown 2015). People that dedicate their free time to the study of organisms and that describe new species without it being part of their professional occupation are often described as amateur taxonomists (Pearson et al. 2011). In some cases they make substantial contributions to our knowledge of organisms. Since they do not have to deal with the administrative side of being a researcher, and do not have to attract funding for their work, they can dedicate virtually all their taxonomy time to doing research. The work presented in this thesis has greatly benefited from the efforts of an extraordinary amateur taxonomist, Sven Olof Ulefors, who dedicates most of his free time to the study of Phoridae (Figure 3). Ulefors has been working with the phorid material from the SMTP since the start of the project in 2003. To date, he has identified about 34,000 specimens to species level; more than 1,200 of these, belonging to close to 1,000 species, have been documented by drawings. The majority of these species are new to science. Largely based on study of this material, Ulefors has identified 58 species groups, a couple of them further divided into subgroups to give a total of 65 groups species groups within Megaselia. Ulefors’ species groups sometimes match Schmitz’s groups, but they often cut across them or represent finer divisions into smaller species groups. In some cases, Ulefors’ tentative species-group concepts were later refined based on preliminary molecular results from the work presented in this thesis.

21

Figure 3: Sven Olof Ulefors.

In paper II, we focus on one of the tentative Megaselia species groups identified by Ulefors, the lucifrons group. We chose a small species group in order to facilitate the development of an effective methodology for carving out and defining natural Megaselia clades. In paper III, we present the first molecular phylogeny of the genus Megaselia and propose the first reasonably comprehensive classification of the genus into species groups. The classification agrees to some extent with Schmitz’s and Ulefors’ morphology-based work, but many of the new species groups cut across the previously recognized groupings. Finally, in paper IV, we treat one of the species groups identified in paper III in more detail, namely the spinigera group. In summary, we describe 47 new species to science in papers II-IV, and provide the first molecularly supported phylogenetic framework for Megaselia, greatly facilitating future taxonomic work on this genus.

22

Material and methods

Material Most of the phorid material studied for this thesis stems from SMTP. SMTP was launched in 2003 with the major goal of making high-quality material of Swedish insects available for the morphological and molecular research needed for a complete national inventory. Up to 75 Malaise traps at 54 localities across Sweden (Figure 4), representing a wide range of habitats, were run for up to three years (for details see Karlsson et al. 2005 and http://www.stationlinne.se/sv/forskning/the-swedish-malaise-trap-project-smtp/). The traps produced about 1,900 samples containing an estimated 80 million specimens. The Diptera and Hymenoptera in particular represent a unique resource for taxonomic and systematics research. To date, more than 500 species previously unknown to science have been discovered in these samples, and new localities in Sweden have been verified for a multitude of additional species.

While the entire SMTP catch is sorted to the level of order, the Diptera and Hymenoptera are sorted into at least family level and in certain cases further to subfamily level or tribes. The material in this thesis is delivered from the SMTP in vials containing phorids only. To date, we have studied more than 44,000 phorid specimens from the SMTP, representing 28 of the 75 traps and 103 of the 1,900 samples. About 34,000 of these specimens were identified to species and some of these specimens were used for subsequent detailed morphological study and molecular sequencing. The remaining specimens are females that cannot be determined currently beyond genus level. In addition to SMTP material, we also studied type specimens from the Schmitz collection at Zoologisches Forschungsmuseum Alexander Koenig, Bonn, Germany (ZFMK) the Lundbeck collection at the Natural History Museum of Denmark (ZMUC) and the phorid collection at the Natural History Museum of Los Angeles County, Los Angeles, USA (LACM).

23

Figure 4: Map of Sweden with dots indicating the Malaise trap sites of the SMTP collecting campaign. Numbers correspond to SMTP trap IDs.

Type specimens of new species, vouchers for sequence data and other important specimens from the SMTP material referenced in the thesis were deposited in the entomology collections at the Swedish Museum of Natural History in Stockholm (NHRS).

Species richness and faunal composition For the estimate of the size and composition of the Swedish insect fauna in paper I, we first divided the known or suspected Swedish insect fauna into

24

families, and the three largest families (Ichneumonidae, Braconidae and Staphylinidae) into subfamilies, producing a total of 636 taxa at the family or subfamily level. The total species richness of the Swedish fauna was then estimated for each of these taxa using two methods. First, we asked experts on each group to make a qualified guess of the size of the total fauna based on their experience. Second, we estimated the total species richness from the SMTP material for eight key taxonomic groups within the Diptera and Hymenoptera using four different species richness estimators. The classic nonparametric estimators Chao1 and Chao2 as well as the capture-mark-recapture (CMR) method have been used and studied widely (O’Hara 2005, Hortal et al. 2006). CMR estimates the number of species from a sample by using the fraction of previously unknown species encountered in the sample, and the total number of known species. The two Chao estimators are based on the number of singletons and doubletons or single or double occurrences. Chao1 is abundance-based and uses the number of singleton and doubleton species, that is, the number of species observed only once or twice in the total sample. Chao2 uses the number of species occurring in only one or two subsamples (in our case traps). Roughly speaking, Chao 1 picks up the increase in diversity expected when extending the sampling period, while Chao 2 focuses on the increase expected when adding more trap sites. As one might expect, Chao2 is more sensitive to grain size (the number of sampling sites or traps) than Chao1 (Hortal et al. 2006). In addition to these estimators, a further nonparametric estimator was developed for paper I, referred to here as the Combined Non-parametric Estimator (CNE). It attempts to combine the strengths of the Chao1 and Chao2 estimators by accounting for the effects of extending both the number of traps and the number of specimens. For details about how to calculate these estimates, see paper I. All of these estimators are nonparametric. Although the performance of nonparametric estimators has been questioned, neither the observed species richness, nor species-area curves or asymptotic parametric estimators seem to perform better (Ulrich and Ollik 2005, Hortal et al. 2006). Therefore, we chose to focus on non-parametric estimators. To study the ecological composition of the Swedish insect fauna, we used the fact that some ecological traits are strongly phylogenetically conserved, so that all members of a taxonomic group used in the analysis would be expected to have the same state of the trait. Thus, we focused on such traits, collected what was known about the state in each group, and then assumed

25

that the same state would apply to the unknown or unstudied species of the group.

Morphological methods Specimen were studied and sorted into morphologically similar groups in 80% ethanol under a stereomicroscope with up to 100X magnification. Whenever a specimen seemed to represent a hitherto unencountered species, it was dissected using an entomological pin and a micropin fixed to a handle. The dissection and further study of the specimen was done in glycerol. Glycerol has a higher viscosity than ethanol, which facilitates dissection since small parts move more slowly and are easier to handle. One hind leg, one wing, the head and the male genitalia were separated from the rest of the body. The dissected specimens were then studied further in a drop of glycerol on a microscope slide with a depression. Studies were focused on characters found in the male genitalia, but additional morphological characters from the thorax, wings and head were examined as well. Pencil drawings were prepared using a camera lucida fitted to a compound light microscope. The pencil drawings were then finalized digitally using Adobe Illustrator before publication. Preparing illustrations as described above of the structurally complex genitalia that are found in the male Megaselia proved to be an effective way to ensure that details could be compared across specimens in a standardized way, using drawings from the same viewing angle. Scanning electronic microscopy (SEM) and digital photography were not used for several reasons. These techniques involve more handling of minute parts, with an increased risk of losing or damaging critical parts of unique specimens. They are also considerably more time consuming without adding much extra information. In the case of SEM, specimen preparation also involves irreversible gold coating, which complicates further study of the specimen parts. With the techniques used here, critical parts can be saved in ethanol or glycerol for future study or sequencing. Based on morphological study of a large number of specimens, we identified 58 groups within Megaselia, a couple of which were further divided into smaller subgroups. These groups and subgroups partly appeared to correspond to groups identified previously by Schmitz (1956) and Disney (1999), but partly represented new groupings and in either case included a large number of new species whose affinities had not been studied previously.

26

Molecular methods For paper III, specimens were selected representing 45 of the 58 species groups that we identified morphologically prior to molecular analyses. The selected species span both Megaselia subgenera and all divisions and rows recognized by Schmitz (paper III, Table 1). Where available we included more than one specimen of each species. For papers II and IV, where we targeted smaller groups within Megaselia, specimens were selected with an as broad geographic range as possible. The molecular markers used in paper II are the nuclear ribosomal D2 region of 28S and the mitochondrial DNA barcode region of COI. A total of 1185 bp were sequenced. In paper III and IV, two additional mitochondrial markers were used, namely 16S and ND1, resulting in a total of 2106 bp. Traditional Sanger sequencing was used.

Data analyses In order to analyze phylogenetic signal in the large amount of data that was generated in this thesis, Bayesian methods were used. Bayesian methods are generally more computationally efficient than Maximum Likelihood (ML) methods and several parameters can be estimated within the same analysis. For instance Bayesian methods can be used to find the best substitution model while running the analysis, whereas a model must be found prior to the analysis when applying ML methods. Details on the programs used and exact settings can be found in each paper.

27

Results

Paper I The results we present in paper I indicate that the Swedish insect fauna is considerably larger than previously thought. By the end of 2012, 26,000 insect species were recorded in DynTaxa, the Swedish national checklist, about 1,300 species more than were known at the start of the SMTP inventory in 2003. However, our combined species richness estimates and expert opinions point to a total insect fauna of roughly 31,000 species, suggesting that 5,000 species still remain to be discovered or officially documented as occurring in Sweden. Poorly known groups (groups where many new species are expected) are almost exclusively found within the Diptera (flies and midges) and Hymenoptera (sawflies, ants, wasps, parasitic wasps). Furthermore, our analysis indicates that most of the missing species are likely to be decomposers and parasitoids, as inferred from the life histories of the taxonomic groups to which they belong. We also detect a geographic trend in the discovery of new species. Specifically, the fauna of southern Sweden seems to have been more completely sampled prior to the SMTP inventory than that of central and northern Sweden. Of the statistical methods we used to estimate total species richness from SMTP data, the simple capture-mark-recapture (CMR) estimator correlated best with expert estimates; in fact, the estimates were remarkably similar. When experts are asked to make an educated guess about the total number of species within their taxonomic groups of expertise, it appears that they tend to think in terms that are similar to the logic of the CMR estimator. However, both the Chao1 and Chao 2 estimators tend to be close to the expert opinions. The CNE estimator, however, indicates much larger species numbers, potentially indicating that experts tend to underestimate the real species diversity. If so, then the true Swedish insect fauna may be considerably larger than 31,000 species.

28

Paper II The study presented in this paper initially focused on the lata group (sensu lato), one of the tentative Megaselia species groups identified during the initial morphology-based survey. Morphologically, the lata group sensu lato is characterized by the combination of a large and spinose labellum, a long costal vein, a long first costal sector, short costal cilia, and a strong proctiger hair. In the aedeagus, the base of the left arm projection is clearly sclerotized. The combination of these features is only found in members of the lata group sensu lato. However, when circumscribed in this way, the group cuts across the subgenera Megaselia and Aphiochaeta recognized by Schmitz, and our preliminary molecular studies indicated that the lata group sensu lato actually consists of two clades that are not very closely related, one falling within the subgenus Megaselia and one within the subgenus Aphiochaeta as defined by Schmitz. We recognized these clades as the lata group sensu stricto (the clade belonging to the subgenus Megaselia) and the lucifrons group (the clade belonging to the subgenus Aphiochaeta), respectively. More detailed morphological studies indicated that there is a difference in the left arm of the aedeagus between the lata group sensu stricto and the lucifrons group. Whereas the left arm projection of the aedeagus is present in the lucifrons group, it is not in the lata group sensu stricto. Furthermore, the presence of a light patch on the hind femur in two of the species in the lucifrons group could possibly be a synapomorphy of the entire lucifrons group, if the absence in M. subnitida is interpreted as a secondary reversal. In the main part of the paper, we focused on the lucifrons group of species. 46 specimens that were identified morphologically belong to the lucifrons group were sequenced for two molecular markers, the D2 region of 28S and COI. The resulting data were analyzed using Bayesian methods of phylogenetic inference. We found three species in the group: Megaselia lucifrons, M. subnitida and a new species, described by us as M. albalucifrons (Figure 5). All three species were well supported by both morphological and molecular characters. M. albalucifrons is so far known from two localities in South and Central Sweden (paper II, Figure 7). M. lucifrons and M. subnitida were considered conspecific by Disney (Disney 1988), but our molecular analyses clearly showed that they are two distinct species. Therefore, we resurrected M. subnitida from synonymy with M. lucifrons.

29

Figure 5: The left side of the epandria in lateral view of Megaselia lucifrons (a), M. subnitida (b) and M. albalucifrons sp. n. (c).

Paper III In this paper, we provide the first comprehensive study of Megaselia relationships based on molecular data. Although basal relationships in the genus remain uncertain, we identify 22 well-supported terminal clades, each containing at least two studied species, and propose that each of these be recognized as an informal species group (Figures 6–7). In addition, we identify nine single-species lineages of uncertain affinities, many of which may turn out to represent distinct higher clades in future analyses. One of these single-species lineages, Myriophora elongata, may well represent a clade corresponding in circumscription to the genus Myriophora as defined by North American authors. At the higher-level, our results suggest (with posterior probability 0.80) that the genus Megaselia consists of a “core Megaselia” clade including 20 of the 22 species groups, and eight of the nine single-species lineages. Outside core Megaselia we find the Myriophora clade, the spinigera group and the ruficornis group (Figure 6). The analysis neither supports nor refutes the monophyly of Megaselia in the broad sense, that is, the group including these three basal clades in addition to core Megaselia. The 22 species groups recognized in the paper generally fall within one of the divisions and rows recognized by Schmitz, suggesting that many of Schmitz’ groups may be natural. However, most of Schmitz’ groups include

a)

b)

c)

0.1 mm

30

more than one of the 22 species groups. Our results neither supported nor strongly refuted the monophyly of the two subgenera Megaselia and Aphiochaeta recognized by Schmitz. However, it is interesting that only two of the 22 species groups – the spinigera group and the minor group – include mixes of species that were or would have been placed in different subgenera by Schmitz. All remaining species groups fall into either one or the other of Schmitz’ subgenera. The 22 species groups correspond less well to the tentative species groups identified by us prior to the analysis, although our groups were smaller and more accurately circumscribed with respect to the studied material than Schmitz’ groups, creating more potential for mismatches. We end the paper by briefly presenting the morphological characteristics of each species group. We also provide molecular and brief morphological descriptions of 45 Megaselia species new to science, represented among the exemplars included in the analysis.

Figure 6: Consensus tree from Bayesian analysis in paper III. Clades with at least 80% PP are cartooned.

0.05

100

64

100

72

100

100

100

80

100

100

68

73

Myriophora elongata

ruficornis group

spinigera group

core Megaselia (165 specimens)

Plectanocnema nudipes

Gymnophora quartomollis

Metopina galeata

Phora holosericea

Borophoga subsultans

Anevrina thoracia

Diplonevra nitidula

Phorinae

Metopininae

Megaselia

31

Paper IV In this paper, we focus on the spinigera group, one of the basal Megaselia groups outside of the core Megaselia clade (see Figure 6). We present a molecular analysis of its phylogenetic position within the subfamily Metopininae, and of relationships within it. The analysis is based on broader sampling of basal Megaselia clades and other related Metopininae lineages than in paper III; it also includes many more specimens belonging to the spinigera group. The phylogenetic analysis supports the recognition of an “extended Megaselia clade” — including Myriophora, Pseudacteon, the M. ruficornis group and core Megaselia — and indicates that the spinigera group may be the sister group of this clade.

Figure 7: Consensus tree of core Megaselia clade from Bayesian analyses in paper III. Clades with more than 80% posterior probability (PP) cartooned.

56100

100

80

10099

65

100

80

99

81

100

64

Megaselia rubella

Megaselia albicaudata

humeralis group

pulicaria complex

minor group

Megaselia flammula

nigra group

lutea group

flava group

pectoralis group

parva group

88glabrifrons group

85cinereifrons group

100manicata group

52

100100

100

Megaselia subpleuralis

nigriceps group

lucifrons group

longiseta group

80

100

100

73

100

54

Megaselia fungivora

pumila group (Kryophiler)

basispinata group

attenuara group

communiformis group

97conformis group

59

52

100

Megaselia hibernans

variana group

0.05

Megaselia levibracchia n. sp.

Megaselia pendulaciliata n. sp.

Anepisternum bare

Anepisternum pilose

Anepisternum pilose + individual setae

Mixed group with bare and pilose anepisternum

32

Furthermore, our analysis shows that the spinigera group contains several species previously placed in Plastophora, Phalacrotophora or Kerophora, and we discuss the possibility of recognizing the spinigera group as a separate genus under one of those names. Unfortunately, there are still uncertainties relating to the placement of the type species of the two oldest of these genera, and we therefore leave the resolution of this question to future studies. We also revise the known Swedish fauna of the group based primarily on the study of SMTP material, and show that the fauna includes a hitherto unknown species, which is described as M. proctoluteipes n. sp. Finally, we present a key to the known Swedish species of the group.

33

Discussion

In biology, there is a long and venerable tradition of estimating species richness. Common methods include analysis of species accumulation curves, parametric fitting of species abundance curves to appropriate models, and non-parametric estimation based on frequencies of rare species (Magurran 2004). A different approach is to ask for the opinions of taxonomic experts working on critical groups, resulting in estimates based on evaluation of a wide range of information sources (Gaston 1991). Despite numerous attempts to estimate global species richness using these methods, the size and composition of the planet’s biota remain elusive. The last decade or so, several studies have converged onto an estimate of approximately 10 M species (Mora et al. 2011; Novotny et al. 2002, references cited therein) but recent DNA-based inventories suggest that this is a significant underestimate (Rodriguez et al. 2013, Tedersoo et al. 2014). Current knowledge of global diversity is simply too scant to allow estimates of total species richness with any degree of accuracy, regardless of method. To obtain a more exact picture of the size and composition of biotas, it is necessary to focus the attention on relatively small geographical regions. The Swedish insect fauna was among the best known in the world even before the STI and SMTP started, and these efforts have ensured that many of the groups that are critical for an accurate insect biodiversity estimate have received recent attention from taxonomists. Of the 636 family-level taxa recognized in our analysis in paper I, only 13 are estimated to contain more than 100 undiscovered species, and most of them are currently worked on by STI- or SMTP-related taxonomic projects. For several of these poorly known groups in the Swedish insect fauna, there is reasonable convergence between expert opinions and the sample-based species richness estimates except for the CNE estimator. There is still considerable uncertainty in our estimate of the size of the Swedish insect fauna, indicated among other things by the significant spread among different species richness estimates. Empirically, it has been observed that undersampling often leads to species richness estimates that are biased downwards (Scharff et al. 2003, O’Hara 2005; Coddington et al. 2009) and we may be seeing such effects in our species richness estimates for some

34

groups. For instance, as many as 26.5% of the Phoridae species occur as singletons in the analyzed data. Theoretically, one might expect both upward and downward biases in the estimators we used. For instance, Chao2 and CNE are based on the assumption that trap sites represent a random sample of possible locations, but SMTP trap sites were deliberately chosen to maximize diversity of the sample, which should result in Chao2 and CNE overestimating true species richness. Furthermore, CNE fails to account for the interaction between the effects of adding years and trap sites to the total sample; some species that would be encountered by adding trap sites might also be added by extending the sampling period. This results in an additional factor leading to CNE overestimating species richness. The large numbers of singletons and doubletons could potentially magnify this latter effect, which might explain a large part of the difference between the CNE estimates and other estimates. Chao1 should be a relatively unbiased estimator of the total fauna one might expect to encounter at the selected trap sites, but it fails to account for the fact that some locally distributed species may be difficult or impossible to collect unless more trap sites are added. Thus, Chao1 clearly represents an underestimate of total species richness. Another potential problem with the SMTP-based estimates is that Malaise traps as a single collecting method were used (Scharff et al. 2003). Although Malaise traps are regarded as an effective method for catching many groups of insects, it is not perfect. Due to practical constraints, however, it was difficult to justify the use of additional trap types in the SMTP inventory. The fact that we used just a single trap type would tend to bias all estimators downwards. Excessive taxonomic splitting could cause the opposite effect. For instance, in a group like Phoridae where male genitalia tend to be species-specific, our experience shows that it is exceedingly rare for molecular analyses to reveal cryptic species but the opposite case occurs regularly. When many species are represented by singletons, it is easy to mistakenly identify morphologically abnormal individuals as distinct species. The combined result of the different effects discussed above is difficult to assess, but it does not seem unreasonable to assume that the CNE estimates generally overestimate species richness, while the other estimates and expert guesses may tend to slightly underestimate species richness. The target groups for which the total species richness was estimated from SMTP data (paper I, Table 1) together represent 1,117 of the 6,300 species estimated by expert to be missing from the known Swedish insect fauna. The CMR, Chao1 and Chao2 estimators all suggest that the expert estimate of 1,117 additional species in these groups is a slight overestimate, while the CNE

35

estimator indicates that the analyzed groups together may contain as much as 4,001 species that remain to be discovered. The CNE estimate would thus suggest a total upper maximum for the Swedish insect fauna of about 50,000 species (24,700 + 6,300 * 4,001 / 1,117). One of the most exciting aspects of the analysis presented in paper I is that it allows us for the first time to answer questions about the composition of insect faunas with some accuracy. However, the exact causes of some of the patterns we observe remain unknown, for instance the difference in number of parasitic species per host species in the phytophagous and saprophagous communities. Ambitious ongoing biodiversity inventories, combined with new molecular techniques, are sure to radically improve our understanding of the size and composition of biotas in the near future. This opens up an entire range of questions, which biologists have not been able to tackle rigorously before. For the first time, biologists can hope to study the structure and function of entire biotas using a whole-systems approach. Even for iconic species that are close to peoples’ hearts and that have been known to man for a long time, we still sometimes have taxonomic issues. The African elephant provides an extreme example: although savanna and forest elephants from across Africa had been considered one species throughout most of the 20th century, recent findings suggest that the two are indeed separate species (Roca et al. 2007). Another example of iconic animals, where phylogenetic relationships are not fully resolved, are the Morpho butterflies of the Neotropics. Even though people are familiar with these large blue butterflies and have been collecting them as souvenirs, their phylogenetic relationships are still unclear and recent studies are still revealing unexpected results (Penz et al. 2012). When only an estimated 10 % of the actual species diversity within a large taxonomic group is known, as is the case for the Phoridae, it is not surprising that virtually nothing is known about their phylogeny or ecological importance. Studying a species’ biology and its role in an ecosystem is a time consuming process, and many years of research are needed in order to merely understand basic facts. Examining phylogenetic relationships can be a way of rapidly charting the major evolutionary lineages, which can help focus field studies and experimental work on relevant groups, and extend existing knowledge to unstudied species. Within the Phoridae and especially within the giant genus Megaselia, we have just begun exploring phylogenetic relationships. Understanding the phylogeny of the large genus Megaselia is the first step towards dividing it into natural subgroups, which will in turn immensely increase the efficiency in the study of its diversity. For instance, this would allow taxonomists to identify manageable projects focused on a limited set of described species in

36

the genus without having to worry about potentially related lineages, thus drastically reducing the time until a more accurate picture of the entire group emerges. In paper III, we take the first steps towards a robust phylogeny-based division of the genus Megaselia into subgroups based on an analysis of four molecular markers. Although many of the relationships within the genus remain uncertain, several interesting results emerge from our analysis. One of the more intriguing is the support for a core Megaselia clade excluding three groups: the spinigera group, the ruficornis group, and the Myriophora clade. The monophyly of core Megaselia is further strengthened by the analysis presented in paper IV, which includes more representatives of basal Megaselia groups and of Megaselia-related genera in the Metopininae. The analysis in paper IV also adds Pseudacteon to the lineages that fall outside but are related to core Megaselia. The availability or potential availability of a number of genus names for these lineages suggest that it may eventually be advisable to recognize all of them as genera distinct from Megaselia, and restrict the latter to the core Megaselia clade. The analyses presented in paper II, paper III and paper IV also offer some of the first insights into the validity of previous taxonomic concepts discussed in the Phoridae. Clearly, although our molecular result contradict some details in Schmitz’ pioneering work on the classification of the group, it appears that some of his results may stand the test of time. Our results may well be compatible with monophyly of one or both of Schmitz’ subgenera, with some modest changes in circumscription, such as the exclusion of species belonging to lineages outside core Megaselia. Many of Schmitz’ divisions and rows may also be sound; our results are inconclusive at this level, but it is noticeable that many of our species groups fall inside only one of Schmitz’ divisions or rows. In contrast, the tentative species groups identified by us prior to molecular analysis appear to be para- or polyphyletic (paper III, Table 1). One reason may be that they are based to a large extent on characters of the male genitalia. While these characters appear to be ideal for separating phorid species, it is possible that they evolve too rapidly to convey much information about relationships above the species level. In contrast, Schmitz’ groups are largely based on external characters, which may not allow species separation but evolve slowly enough to retain important signal about higher relationships within the genus. Even though much of the progress in phorid taxonomy presented in this thesis is based on molecular analyses, we believe that it is important to continue studying phorid morphology. For instance, morphological traits

37

provide a cheaper, faster and simpler method of identification than DNA analyses, so knowledge of diagnostic morphological characters is still of considerable practical significance. Male genitalia in Megaselia appear to provide a very reliable means of species identification; in fact, male genitalia appear to contain as much information relevant for species delimitation as molecular markers. Over several years of molecular analyses of phorids, we have failed to detect a single species that was not revealed already by examination of male genitalic characters. This contrasts with many other groups of insect, e.g. parasitic wasps of the family Braconidae (Stigenberg and Ronquist 2011), where male genitalia are less complex (or less well studied) and molecular studies tend to reveal a high proportion of cryptic species.

38

Conclusion

Within this thesis I first explored the size and composition of the Swedish insect fauna in paper I. In the following papers (II-IV) I concentrated on one of the most poorly known groups within the Swedish Diptera fauna, the family Phoridae (scuttle flies). Specifically, I contributed to the knowledge of the Swedish phorid fauna and the phylogeny and classification of the mega-diverse phorid genus Megaselia. In total, 47 new species of Megaselia are described in the papers within this thesis and the first molecular phylogeny of the genus Megaselia is presented. Future projects stemming from this thesis should explore the use of genomic data for phylogenetic analyses of the genus Megaselia. The use of more data would certainly be beneficial in order to untangle the relationships of species within this large radiation. The analyses presented in papers III-IV leave much of the basal splits within the core Megaselia clade unresolved. Furthermore, the species groups recognized in paper III should be further studied and a broader material from a wider geographic range should be sampled and analyzed. With a step-by-step approach to refining our phylogenetic knowledge about the radiation, the giant genus Megaselia could be divided into successively finer and more accurately circumscribed clades, increasingly opening up the group to a range of more detailed studies of subgroups of manageable size. Eventually, it seems likely that several of these clades will be recognized as genera distinct from Megaselia. Charting biodiversity may be the single most important field of research in order to understand the complexity of ecosystems and the world we live in. In the long run, the solution to our problems of overpopulating planet Earth may be found here. There may well be a species of plant that we have not found yet that could turn out to be a superior source of food, clothing or building material. There may well be a species of parasitic wasp that – provided we do not let it go extinct – may control a pest species of flies that otherwise would have destroyed our harvests in the year 2097. We can only speculate about what there is to be found in the world around us. What we do know is that habitats are being destroyed at a rate that cannot be recovered by nature alone and that species are going extinct at a higher rate than what we would consider normal. Which these doomed species are, we

39

often do not know. What we do know is that describing species takes time, and studying their life histories and ecology is even more time-consuming. We have been doing it for about 300 years and we are not even close to finishing the job. This thesis, focused on one of the most abundant and ubiquitous insect groups, is offered as one small contribution towards the urgent task of completing the inventory of life on earth. Maybe it will only serve to satisfy our curiosity, maybe it will turn out to represent a small but important step towards saving humankind.

40

Svensk Sammanfattning

Den här avhandlingen handlar om insektsdiversitet i Sverige i allmänhet och diversiteten hos en stor grupp av små flugor – puckelflugor – i synnerhet. Det pratas nu för tiden mycket om att vi måste bevara den biologiska mångfalden. Genom förstörelse av habitat riskerar många arter att dö ut medan det för många andra redan är för sent. Samtidigt är vi långt ifrån att känna till alla arter på denna planet. Den första artikeln i avhandlingen använder ett par olika metoder och nya data för att försöka uppskatta hur många insektsarter som totalt finns i Sverige. Bland insekterna finns dels välkända grupper som fjärilar och skalbaggar, där vi i princip känner till varenda art som finns i vårt land, dels grupper som tvåvingar (flugor och myggor) och steklar (med bl. a. myror, getingar, bin, men framförallt parasitsteklar), där mycket av mångfalden fortfarande är okänd. Det är svårt att föreställa sig att det fortfarande finns mycket okänt i ett land som Sverige, där traditionen att beskriva arter och studera mångfalden går ända tillbaka till Carl von Linné, men faktum är att det fortfarande finns många nya arter kvar att upptäcka. Sedan ett tiotal år finns ett statligt finansierat projekt, Svenska Artprojektet, som har som mål att kartlägga alla flercelliga organismer i Sverige (http://www.artdatabanken.se/verksamhet-och-uppdrag/svenska-artprojektet/). Med finansiering av Svenska Artprojektet startades 2003 Svenska Malaisefälleprojektet eller SMTP (http://www.stationlinne.se/sv/forskning/the-swedish-malaise-trap-project-smtp/), den första systematiska inventeringen av den svenska insektsfaunan. Med Malaisefällor samlades uppskattningsvis 80 miljoner insekter in från hela Sverige under tre års tid, och materialet har sedan dess successivt sorterats i mindre taxonomiska grupper för att göra det så tillgängligt för taxonomer som möjligt. Med hjälp av data från SMTP har vi uppskattat det totala antalet arter i åtta utvalda insektsgrupper genom att använda välkända metoder. Dessa metoder tar i beaktning vilka arter som är kända i det studerade området (i det här fallet Sverige), samt vilka arter som faktiskt har hittats och i vilket antal. Speciellt viktiga är här arter som bara hittas i en eller två exemplar, eftersom frekvensen av dessa ger en indikation på hur komplett det studerade

41

underlaget är. Ju fler arter som hittas i bara en eller två exemplar, desto längre ifrån det totala antalet arter är insamlingen. Som komplement till dessa uppskattningar bad vi experter att bedöma hur stor den svenska insektsfaunan är. Totalt omfattar faunan 636 grupper, om man delar in den i familjer och de tre största familjerna i underfamiljer. Experterna är ofta professionella taxonomer men kan också vara personer som har systematisk forskning som fritidsintresse. De senare kallar vi ofta för amatörtaxonomer trots att de oftast är väl så insatta i sin grupp som professionella systematiker. Till experterna ställde jag frågan ”Hur många arter tror du att det finns totalt i din grupp i Sverige?” Även om svaret kan te sig som en vild gissning, får vi inte glömma att det baseras på många års erfarenhet och en bra känsla för gruppen. Dessutom har många av experterna tack vare Svenska artprojektet kunnat specialstudera den svenska faunan under senare år. Det visade sig också när vi jämförde expertgissningarna med våra beräknade siffror (i de få grupper där vi hade en sådan siffra) att experterna alltid var nära. Sammanlagt tror experterna att vi har över 31 000 arter i Sverige, att jämföra med de nu kända 26 000 arterna. Vår analys indikerar att det bara är 13 av våra 636 insektsgrupper där vi kan förvänta oss att upptäcka mer än 100 nya arter och där stora forskningsinsatser är nödvändiga. En av dessa grupper är tvåvingefamiljen puckelflugor. I Sverige är idag 404 arter av puckelflugor rapporterade i Dyntaxa, den officiella svenska checklistan (http://dyntaxa.se). Hela 278 arter ingår i ett enda släkte – Megaselia. Sven Olof Ulefors uppskattar dock det totala artantalet i Sverige till 1 100. Globalt har släktet Megaselia över 1 600 arter (http://phorid.net/pcat/numSpecPerGen.php), vilket gör det enormt stort och svåröverskådligt. Innan jag började med denna avhandling hade inga försök gjorts att dela upp släktet i mindre grupper med hjälp av molekylära metoder. Däremot fanns ett arbete där en indelning på morfologisk grund hade påbörjats, men tyvärr inte avslutats (Schmitz 1951, 1955, 1956, 1957, 1958, Schmitz and Beyer 1965a, 1965b, Schmitz et al. 1974, Schmitz and Delage 1981). Med hjälp av detta arbete samt egna morfologiska studier hade Sven Olof Ulefors, en amatörtaxonom som är expert på puckelflugor och speciellt intresserad av Megaselia, kommit fram till en indelning av släktet i totalt 58 grupper, varav ett par delats in ytterligare i delgrupper. För att testa den här morfologiska indelningen, men även för att få ett så brett material för en fylogenetisk studie av släktet Megaselia som möjligt, samlade jag ihop så många arter puckelflugor som möjligt med målet att få

42

med minst två arter från varje grupp, för större grupper ännu fler. Jag lyckades att få med individer från 45 grupper och analyserade totalt 175 individer i 145 arter av Megaselia med molekylära metoder. För detta arbete sekvenserade jag fyra olika gener och analyserade sedan datasetet med Bayesianska metoder. Resultatet är den första molekylära fylogenin (släktskapsträdet) över släktet Megaselia, vilken presenteras i artikel III i avhandlingen. Baserat på fylogenin föreslår vi en indelning av släktet i 22 monofyletiska artgrupper som omfattar två eller fler arter. En monofyletisk grupp är en grupp arter som delar en gemensam förfader de inte delar med andra arter. Artgrupperna stämmer bara delvis överens med de grupper som identifierats tidigare av Schmitz och Ulefors. Morfologiska egenskaper för alla artgrupper beskrivs i artikeln. Ytterligare åtta arter uppträder ensamma i trädet, och mer omfattande analyser behövs för att avgöra om de representerar egna artgrupper eller om de bäst förs till en av de artgrupper vi definierar i artikeln. Ytterligare analyser kommer att vara nödvändiga för att få en fullständig upplösning av Megaselia-trädet, men som ett redskap för vidare taxonomiskt och fylogenetiskt arbete så kommer den här artikeln att kunna användas länge. Förutom det fylogenetiska trädet som presenteras, så beskrivs även 45 arter som nya för vetenskapen. I de resterande två artiklarna (II och IV) studerar vi två av artgrupperna som vi identifierat i artikel III, lucifrons-gruppen och spinigera-gruppen. Bägge grupperna analyseras med både molekylära och morfologiska metoder baserat huvudsakligen på material från SMTP. I vardera artikeln beskrivs en ny art för vetenskapen. Sammanlagt har forskningen i denna avhandling bidragit till vår kunskap om den svenska insektsfaunan dels genom att peka ut de taxonomiska grupper där vi kan förvänta oss att hitta många nya arter och dels genom att föra forskningen i en av dessa grupper framåt. Genom att presentera den första molekylära fylogenin för det artrika släktet Megaselia har vi lagt grunden för att kunna utforska denna grupp på ett betydligt effektivare och mer strukturerat sätt än tidigare.

43

Acknowledgements

Among many other things I have learned while writing my thesis, for me one of the most important ones is this: Even though the cover of my PhD thesis bears my name only, I wouldn’t have been able to put it together without the support of many others! Any attempt to list every person in any way involved into this work is bound to fail; I know that. Therefore, if you found this page: Thank you! Thank you for kind words, difficult questions, a cup of coffee, lending me a book or for at some point leading me into the right direction! However, I still want to thank quite a few people specifically and I want to start with thanking my supervisors and the members of my follow-up group. Fredrik, thank you for letting me do my PhD project with you. I still remember the excitement when I found out that I was accepted for doing my PhD in your department. Kjell Arne, thank you for your help and encouragement during my time at the entomology and later zoology department. Ulf and Bertil, thank you for valuable and relevant comments and discussions during my follow-up meetings. Sven Olof, you have been a wonderful mentor to me. You are incredibly generous in sharing your knowledge about phorids in general and Megaselia in particular with me. Without you, I wouldn’t have known where to even begin. Mattias, you are a walking encyclopedia of everything and I have had the pleasure to share my office and many, many interesting conversations with you (Fredrik, don’t look too closely here, we did talk biology quite often). Kevin, I still think it’s sad that you are not my officemate anymore; you always had a kind word for me. You still do. And you know what? I still consider you a fellow Dipterist. Dave and Kajsa, thank you for letting me hang out with the SMTP group back in the days when I was on Öland doing plant ecology. I am grateful that you offered me a job sorting insects with you and introducing me to Fredrik. I had a great time working together with Pelle, Svante, Annie, Erika and Jon and it was always great fun to visit you again at Station Linné. I also met Julia there who started her own PhD project (and finished way earlier)

44

together with me. You continue to be a truly appreciated friend both in and out of work. Also, you named a wasp after me! Thank you! At the entomology department at NRM I want to thank all my colleagues for making our work environment such a friendly place! I loved being a part of the department and I hope you will let me visit. Special thanks to Hege, Gunvi and Niklas for all your help with practical stuff and an open ear when needed! Johannes, I could count on you for advice when I needed it, thank you for that! Thank you, Yngve for your moral support and friendly conversations. Marie, Jane and Anna-Lena, thank you for never losing your patience with countless questions about budgets and other practical aspects of being a PhD student at the museum! Pia, Martin, Bodil and Keyvan, thank you all for your help and support with my lab work at MSL! I also want to give a big thank you and a huge hug to all my PhD student colleagues at entomology! Emma (an extra thank you for tracking down that primer sequence last minute!), James, Jeanette, Jonas, Julia, Marianne, Rasa (thank you for the soap!), Rasmus (I missed being an actual PhD student colleague by a few months, I think), Tobias K. (even though you’re not in entomology) and Tobias M. – I discussed both research-related topics and life as a PhD student with all of you. Thank you! Julia, Rasa and Tobias M., I want to especially thank you for your friendship during a difficult time, I appreciate that immensely. Marianne and Tobias M., you were already experienced PhD students when I joined the department and helped me not only in the lab but also to understand the special challenges with being placed at the museum while employed by the university. You also introduced me to teaching courses in physiology and zoology that were a bit out of my comfort zone in the beginning but that I learned to enjoy teaching. Thank you, Bertil, Heiner and Micke for having me as an assistant on your courses and listening to my ideas. I also want to mention all my other course assistant colleagues (Lina E., Lina S., Lina T., Jeannette, Sandra, Timm, Piotr, Bea, Souli) and thank you for enjoyable teaching and a shared fear for the “visningsmikroskop”. Thank you, Erik for all technical and moral support during these courses. I also want to say thank you to Kristin. We have not only shared teaching experiences but also realized that some of the same things happen to both of us. I will never forget teaching faunistics HT13 with you and me not being able (or allowed) to carry anything heavy. Erland, you started it all (the teaching) standing in my office one day in early 2009 asking me to be your course assistant. You once said that you never regretted that. I don’t either. Teaching faunistics I never had to learn to enjoy, I just did. I taught together with Magne, Martin O., Martin B., Lina S., Julia, Rasmus, Jonas, Robert, Kristin and Alex and I had lots of fun.

45

Also, Alex, thank you for reminding me of some of the classics of German music! At this point I would also like to thank all of my students who have made teaching unpredictable, time-consuming and stressful. I have learned tremendously and it was a true joy! Anette, Berit and Siw, thank you for never losing patience with any of many questions and requests regarding everything administrative at Stockholm University! Mama und Papa, Laura und Justus: Ich bin, wer ich bin – euretwegen. Ohne Eure Unterstützung, Erziehung und Liebe wäre ich heute nicht hier. Danke! Johan, vilken resa vi har gjort! Jag önskar oss att vi fortsätter att växa tillsammans. Jag älskar dig. Wilma und Anton, Ihr seid das Beste, das mir je passiert ist. Ihr zeigt mir täglich, was wichtig ist. Lastly, I would like to thank Fredrik again. You have been incredibly patient with me but also encouraging and pushing me forwards when you found the need to do so. I sincerely wish for all PhD students out there to feel supported by their supervisors as I did. I know you cared.

46

References

Agnarsson I, Kuntner M (2007) Taxonomy in a changing world: seeking solutions for a science in crisis. Systematic biology 56: 531–9. doi: 10.1080/10635150701424546

Azevedo GHF, Faleiro BT, Magalhães ILF, Benedetti a. R, Oliveira U, Pena-Barbosa JPP, Santos MTT, Vilela PF, de Maria M, Santos a J (2013) Effectiveness of sampling methods and further sampling for accessing spider diversity: a case study in a Brazilian Atlantic rainforest fragment. Insect Conservation and Diversity: 1–11. doi: 10.1111/icad.12061

Bonet J, Ulefors S-O, Viklund B, Pape T (2011) Species richness estimations of the megadiverse scuttle fly genus Megaselia (Diptera: Phoridae) in a wildfire-affected hemiboreal forest. Insect Science 18: 325–348. doi: 10.1111/j.1744-7917.2010.01362.x

Brown BV (2005) Malaise trap catches and the crisis in Neotropical Dipterology. American Entomologist 51: 180–183. Available from: http://www.ingentaconnect.com/content/esa/ae/2005/00000051/00000003/art00011 (May 22, 2012).

Brown BV, Kung G-A, Porras W (2015) A new type of ant-decapitation in the Phoridae (Insecta: Diptera). Biodiversity Data Journal 3: e4299. doi: 10.3897/BDJ.3.e4299

Brown BV., Smith PT (2010) The bee-killing flies, genus Melaloncha Brues (Diptera: Phoridae): a combined molecular and morphological phylogeny. Systematic Entomology 35: 649–657. doi: 10.1111/j.1365-3113.2010.00540.x

Ceballos G, Ehrlich PR, Barnosky AD, García A, Pringle RM, Palmer TM (2015) Accelerated modern human – induced species losses: entering the sixth mass extinction. Sciences Advances 1: 1–5. doi: 10.1126/sciadv.1400253

Coddington JA, Agnarsson I, Miller JA, Kuntner M, Hormiga G (2009) Undersampling bias: the null hypothesis for singleton species in tropical arthropod surveys. JOURNAL OF ANIMAL ECOLOGY 78: 573–584. doi: 10.1111/j.1365-2656.2009.01525.x

Cook CE, Mostovski M (2002) 16S mitochondrial sequences associate morphologically dissimilar males and females of the family Phoridae (Diptera). Biological Journal of the Linnean Society 77: 267–273.

47

Available from: http://onlinelibrary.wiley.com/doi/10.1046/j.1095-8312.2002.00109.x/full (May 22, 2012).

Disney RHL (1988) Three new species of Megaselia (Dipt., Phoridae) from the British Isles, plus new synonyms. Entomologist’s Monthly Magazine 124: 7–14.

Disney RHL (1999) A troublesome sibling species complex of scuttle flies ( Diptera : Phoridae ) revisited. Journal of natural history 33: 1159–1216. Available from: http://www.tandfonline.com/doi/abs/10.1080/002229399299987 (May 15, 2012).

Dunn RR (2005) Modern Insect Extinctions, the Neglected Majority. Conservation Biology 19: 1030–1036. doi: 10.1111/j.1523-1739.2005.00078.x

Fontaine B, van Achterberg K, Alonso-Zarazaga MA, Araujo R, Asche M, Aspöck H, Aspöck U, Audisio P, Aukema B, Bailly N, Balsamo M, Bank R a., Belfiore C, Bogdanowicz W, Boxshall G, Burckhardt D, Chylarecki P, Deharveng L, Dubois A, Enghoff H, Fochetti R, Fontaine C, Gargominy O, Lopez MSG, Goujet D, Harvey MS, Heller K-G, van Helsdingen P, Hoch H, De Jong Y, Karsholt O, Los W, Magowski W, Massard J a., McInnes SJ, Mendes LF, Mey E, Michelsen V, Minelli A, Nafrıa JMN, van Nieukerken EJ, Pape T, De Prins W, Ramos M, Ricci C, Roselaar C, Rota E, Segers H, Timm T, van Tol J, Bouchet P (2012) New Species in the Old World: Europe as a Frontier in Biodiversity Exploration, a Test Bed for 21st Century Taxonomy Schierwater B (Ed). PLoS ONE 7: e36881. doi: 10.1371/journal.pone.0036881

Frodin DG (2004) History and concepts of big plant genera. Taxon 53: 753–776. doi: 10.2307/4135449

Frøslev TG, Jeppesen TS, Læssøe T, Kjøller R (2007) Molecular phylogenetics and delimitation of species in Cortinarius section Calochroi (Basidiomycota, Agaricales) in Europe. Molecular Phylogenetics and Evolution 44: 217–227. doi: 10.1016/j.ympev.2006.11.013

Gaston KJ (1991) The Magnitude of Global Insect Species Richness. Conservation Biology 5: 283–296.

Gillman LN, Wright SD, Cusens J, Mcbride PD, Malhi Y, Whittaker RJ (2015) Latitude, productivity and species richness. Global Ecology and Biogeography 24: 107–117. doi: 10.1111/geb.12245

Hash JM, Brown BV (2015) Revision of the New World Species of the Millipede-Parasitic Genus Myriophora Brown (Diptera: Phoridae). Zootaxa 4035: 1–79.

Hash JM, Brown BV, Smith PT (2011) Preliminary use of DNA sequences for Dohrniphora (Diptera: Phoridae) phylogeny and taxonomy.

48

Zootaxa 2991: 13–20. Hillebrand H (2004) On the Generality of the Latitudinal Diversity Gradient.

The American Naturalist 163: 192–211. Hortal J, Borges PA V, Gaspar C (2006) Evaluating the performance of

species richness estimators: sensitivity to sample grain size. Journal of Animal Ecology 75: 274–287. doi: 10.1111/j.1365-2656.2006.01048.x

Jendek E, Poláková J (2014) Host Plants of World Agrilus (Coleoptera, Buprestidae). Springer, Ottawa, 1-679 pp. doi: 10.1007/978-3-319-08410-7

Joppa LN, Roberts DL, Myers N, Pimm SL (2011) From the Cover: Biodiversity hotspots house most undiscovered plant species. Proceedings of the National Academy of Sciences of the United States of America 108: 13171–13176. doi: 10.1073/pnas.1109389108

Karlsson D, Pape T, Johanson KA, Liljeblad J (2005) Svenska Malaisefälleprojektet, eller hur många arter steklar, flugor och myggor finns i Sverige? Entomologisk tidskrift 126: 43–53.

Lamarre GPA, Molto Q, Fine PVA, Baraloto C (2012) A comparison of two common flight interception traps to survey tropical arthropods. ZooKeys 216: 43–55. doi: 10.3897/zookeys.216.3332

Magurran AE (2004) Measuring Biological Diversity. Blackwell Publishing, Oxford.

Miller GH, Magee J, Smith M, Spooner N, Baynes A, Lehman S, Fogel M, Johnston H, Williams D, Clark P, Florian C, Holst R, DeVogel S (2016) Human predation contributed to the extinction of the Australian megafaunal bird Genyornis newtoni ~47 ka. Nature Communications 7: 1–7. doi: 10.1038/ncomms10496

Mora C, Tittensor DP, Adl S, Simpson AGB, Worm B (2011) How many species are there on Earth and in the ocean? PLoS biology 9: e1001127. doi: 10.1371/journal.pbio.1001127

Novotny V, Basset Y, Miller SE, Weiblen GD, Bremer B, Cizek L, Drozd P (2002) Low host specificity of herbivorous insects in a tropical forest. Nature 416: 841–844. doi: 10.1038/416841a

O’Hara RB (2005) Species richness estimators: how many species can dance on the head of a pin? Journal of Animal Ecology 74: 375–386. doi: 10.1111/j.1365-2656.2005.00940.x

Pearson DL, Hamilton AL, Erwin TL (2011) Recovery Plan for the Endangered Taxonomy Profession. BioScience 61: 58–63. doi: 10.1525/bio.2011.61.1.11

Penz CM, Devries PJ, Wahlberg N (2012) Diversification of Morpho butterflies (Lepidoptera, Nymphalidae): A re-evaluation of morphological characters and new insight from DNA sequence data. Systematic Entomology 37: 670–685. doi: 10.1111/j.1365-

49

3113.2012.00636.x Régnier C, Achaz G, Lambert A, Cowie RH, Bouchet P, Fontaine B (2015)

Mass extinction in poorly known taxa. Proceedings of the National Academy of Sciences 112: 201502350. doi: 10.1073/pnas.1502350112

Roca AL, Georgiadis N, O’Brien SJ (2007) Cyto-nuclear genomic dissociation and the African elephant species question. Quaternary International 169–170: 4–16. doi: 10.1016/j.quaint.2006.08.008

Rodriguez JJ, Fernández-Triana JL, Smith MA, Janzen DH, Hallwachs W, Erwin TL, Whitfield JB (2013) Extrapolations from field studies and known faunas converge on dramatically increased estimates of global microgastrine parasitoid wasp species richness (Hymenoptera: Braconidae). Insect Conservation and Diversity 6: 530–536. doi: 10.1111/icad.12003

Scharff N, Coddington JA, Griswold CE, Hormiga G, De P (2003) When to quit? Estimating spider species richness in a northern European deciduous forest. The Journal of Arachnology 31: 246–273.

Scheffers BR, Joppa LN, Pimm SL, Laurance WF (2012) What we know and don’t know about Earth’s missing biodiversity. Trends in Ecology and Evolution 27: 501–510. doi: 10.1016/j.tree.2012.05.008

Schmitz H (1926) Klassifikation der Phoriden und Gattungsschlüssel. Natuurhistorisch Maandblad 15: 9–12.

Schmitz H (1927) Klassifikation der Phoriden und Gattungsschlüssel. Natuurhistorisch Maandblad 16: 1–2.

Schmitz H (1956) 33. Phoridae. In: Lindner E (Ed), Die Fliegen der palaearktischen Region. E. Schweizerbart’sche Verlagsbuchhanglung (Nägele u. Obermiller), Stuttgart, 369–416.

Shaw MR, Hochberg ME (2001) The neglect of parasitic Hymenoptera in insect conservation strategies : The British fauna as a prime example. Journal of Insect Conservation 5: 253–263.

Smith PT, Brown BV (2010) A molecular phylogenetic analysis of genus Anevrina ( Diptera : Phoridae ), with the description of a new species and updated world key. Zootaxa 2397: 29–40.

Soon V, Kesküla T, Kurina O (2011) Strepsiptera species in Estonia. Entomologica fennica 22: 213–218.

Steffen W, Crutzen J, McNeill JR (2007) The Anthropocene: are humans now overwhelming the great forces of Nature? Ambio 36: 614–621. doi: 10.1579/0044-7447(2007)36[614:TAAHNO]2.0.CO;2

Stigenberg J, Ronquist F (2011) Revision of the Western Palearctic Meteorini (Hymenoptera, Braconidae), with a molecular characterization of hidden Fennoscandian species diversity. Zootaxa 3084: 1–95.

50

Tedersoo L, Bahram M, Põlme S, Kõljalg U, Yorou NS, Wijesundera R, Ruiz LV, Vasco-Palacios AM, Thu PQ, Suija A, Smith ME, Sharp C, Saluveer E, Saitta A, Rosas M, Riit T, Ratkowsky D, Pritsch K, Põldmaa K, Piepenbring M, Phosri C, Peterson M, Parts K, Pärtel K, Otsing E, Nouhra E, Njouonkou AL, Nilsson RH, Morgado LN, Mayor J, May TW, Majuakim L, Lodge DJ, Lee SS, Larsson K-H, Kohout P, Hosaka K, Hiiesalu I, Henkel TW, Harend H, Guo L, Greslebin A, Grelet G, Geml J, Gates G, Dunstan W, Dunk C, Drenkhan R, Dearnaley J, Kesel A De, Dang T, Chen X, Buegger F, Brearley FQ, Bonito G, Anslan S, Abell S, Abarenkov K (2014) Global diversity and geography of soil fungi. Science 346. doi: 10.1126/science.aaa1185

Tomasson K, Tammaru T, Kurina O (2014) Harvestmen ( Arachnida : Opiliones ) in Estonia : results of the Estonian Malaise Trap Project. Entomologica Fennica 10: 142–156.

Ulrich W, Ollik M (2005) Limits to the estimation of species richness: The use of relative abundance distributions. Diversity and distributions 11: 265–273. doi: 10.1111/j.1366-9516.2005.00127.x

Vårdal H, Taeger A (2011) The life of René Malaise: From the wild east to a sunken island. Zootaxa 52: 38–52.

Weinmann D, Disney RHL (1997) Two new species of Phoridae (Diptera) whose larvae associate with large spiders (Araneae: Theraphosidae). Journal of Zoology 243: 319–328.