Chapter 1: General Introduction

TABANIDAE AND CULICIDAE IN THE NORTHERN BOREAL REGION OF

ONTARIO

A Thesis Submitted to the Committee on Graduate Studies

in Partial Fulfillment of the Requirements for the Degree of Master of Science

in the Faculty of Arts and Science

TRENT UNIVERSITY

Peterborough, Ontario, Canada

Environmental and Life Sciences Graduate Program M.Sc

January 2015

ABSTRACT

I studied the abundance, distribution and diversity of horse fly and deer fly species (Diptera: Tabanidae) and mosquito species (Diptera: Culicidae) in the boreal forest region of northern Ontario in 2011 and 2012. I collected 19 mosquito species, including one species new for Ontario, Aedes pullatus (Coquillett). I documented 11 northern and one southern range extension. I also collected a total of 30 species of horse and deer flies, including one new species of horse fly for Ontario, Hybomitra osburni

(Hine). Results were inconsistent with a hypothesis of colonization of dipteran species from west to east. I examined the trapping biases of Malaise and sweep sampling for horse and deer flies and found that Malaise traps collected fewer individuals than sweep netting (850 versus 1318) but more species (28 versus 22). Consequently, I determined that surveys of diversity benefit from the use of multiple trapping methods. I also examined how blood-feeding (anautogeny) requirements affect the distribution patterns of Tabanidae. Ultimately, there are likely multiple factors that affect the expression of anautogeny in Tabanidae.

Keywords: Tabanidae, Culicidae, Northern Ontario, Hudson Bay Lowlands, Species distribution, Autogeny, Boreal Forest, Anautogeny

ACKNOWLEDGEMENTS

I wish to thank the Ontario Ministry of Natural Resources Northeast Science and

Information Section and Wildlife Research & Development Section for coordinating my field work and allowing me to take part in the Far North Biodiversity Project. I would also like to thank the Far North Branch for providing funding. I wish to express my thanks to the Northern Scientific Training Program (NSTP) for providing me with travel support.

I am incredibly grateful to the members of the Far North Biodiversity Project for helping me collect my samples: Dean Phoenix, Todd Copeland, Jane Devlin, Kevin

Downing, Derek Goertz, Alex Howard, Ian Fife and Shannon Page. I am also grateful to the countless number of volunteers of the project that helped with sample collection.

I want to thank my co-supervisors David Beresford and Ken Abraham for their constant support, guidance and encouragement. I would also like to thank my committee member, Erica Nol, for her feedback and guidance.

I also want to thank my fellow lab mates, Donald Bourne, Sarah Langer, Scott

Larkin and Marco Raponi for their unwavering support.

I especially want to thank my parents and family for their constant support and encouragement. I owe the biggest thanks to Jen, who was my biggest supporter through it all.

Last but not least, I sincerely thank the First Nations communities,

Kitchenuhmaykoosib Inninuwug, Keewaywin, and Fort Albany for their hospitality.

III

TABLE OF CONTENTS

ABSTRACT ...... II

ACKNOWLEDGEMENTS ...... III

LIST OF FIGURES ...... VI

LIST OF TABLES ...... VII

LIST OF ABBREVIATIONS ...... VIII

CHAPTER 1: GENERAL INTRODUCTION ...... 1 CHAPTER 2: NEW RANGE RECORDS OF MOSQUITO SPECIES FROM NORTHERN ONTARIO ...... 5

ABSTRACT ...... 5

INTRODUCTION ...... 5

MATERIALS AND METHODS ...... 7

Analysis...... 8

RESULTS ...... 10

New Ontario record...... 10

Northward range extensions ...... 11

Southward range extensions ...... 12

Range gap infills ...... 13

DISCUSSION ...... 14 CHAPTER 3: NEW RANGE RECORDS, AND A COMPARISON OF SWEEP NETTING AND MALAISE TRAP CATCHES OF HORSE FLIES AND DEER FLIES (DIPTERA: TABANIDAE) IN NORTHERN ONTARIO...... 23

ABSTRACT ...... 23

INTRODUCTION ...... 23

MATERIALS AND METHODS ...... 25

Analysis...... 26

RESULTS ...... 27

New Ontario record...... 27

Range extensions ...... 28

Gap infills...... 28

Trap comparison ...... 29

DISCUSSION ...... 30

CHAPTER 4: ANAUTOGENY IN TABANIDAE ...... 38

ABSTRACT ...... 38

INTRODUCTION ...... 38

Rationale for hypotheses ...... 41

Predictions...... 43

MATERIALS AND METHODS ...... 45

Sampling ...... 45

Species Classification ...... 45

Analysis...... 45

RESULTS ...... 46

DISCUSSION ...... 47

CHAPTER 5: GENERAL DISCUSSION ...... 55

LITERATURE CITED ...... 60

APPENDICES ...... 68

APPENDIX 1: SITE DESCRIPTIONS ...... 68

APPENDIX 2: SITE PICTURES ...... 70

LIST OF FIGURES

FIGURE 2.1. Rarefaction analysis of mosquito catches (means and SDs) within 150 km of Big Trout Lake and Sandy Lake in 2011 (closed circles) and within 150 km of Ft. Albany in 2012 (open circles). The inset map of Ontario shows the sampling locations in 2011 and 2012...... 17

FIGURE 2.2. Fitted lognormal distributions of mosquito catches within 150 km of Big Trout Lake and Sandy Lake in 2011 and within 150 km of Ft. Albany in 2012. The area of the region left of the veil line represents species that were too rare to be sampled with our methodology...... 18

FIGURE 3.1. Rarefaction analysis showing the expected number of Tabanidae species (y axis) for smaller total catch sizes (x axis), for 2011 (closed circles) and 2012 (open circles). The inset map shows sample locations in both years. Error bars represent standard deviations...... 32

FIGURE 3.2. Rarefaction analysis of 2011 and 2012 Tabanidae collection data, separated by trapping method. Malaise traps (circles) and sweep netting (diamonds) in 2011 (closed) and 2012 (open). Error bars represent standard deviations...... 33

FIGURE 4.1. Sweep and Malaise catches of anautogenous (open circles and dashed trendline) and autogenous (solid circles and solid trendline) Tabanidae species from 2011 in the western boreal forest of Ontario...... 51

FIGURE 4.2. Sweep and Malaise catches of anautogenous (open circles and dashed trendline) and autogenous (solid circles and solid trendline) Tabanidae species from 2012 in the eastern boreal forest of Ontario...... 52

APPENDIX FIGURE 1. Typical Malaise trap setup in a Fen bordered by a black spruce (Picea mariana) forest...... 70

APPENDIX FIGURE 2. Typical camp setup showing malaise trap placement in a fen bordered by a black spruce (Picea mariana) forest...... 71

APPENDIX FIGURE 3. Typical camp setup in a bog. Malaise trap was placed behind the photographer...... 71

APPENDIX FIGURE 4. Typical camp setup in a bog bordered by a black spruce (Picea mariana) forest...... 72

LIST OF TABLES

TABLE 2.1. Culicidae species collected in 2011 within 150 km of Big Trout Lake and Sandy Lake, and in 2012 within 150 km of Fort Albany...... 19

TABLE 2.2. Culicidae species found at each sampling site. Dates indicate when sampling was conducted. There are only 11 sample sites listed in 2011 because collections from one of the July 12–21 sites were damaged by a bear. Reference numbers (TUIC) refer to the voucher specimens of the Trent University Insect Collection, housed at Trent University...... 21

TABLE 3.1. Species and number of Tabanidae collected in 2011 and 2012 using Malaise traps and sweep netting, with abundance records...... 34

TABLE 3.2. Tabanidae species listed for each sampling location and date. Only 11 sampling sites are included for 2011 as one additional planned sampling site was unable to be surveyed due to weather...... 36

TABLE 4.1. Classification of autogenous and anautogenous Tabanidae species collected in 2011 and 2012 in the Far North of Ontario. Species are grouped based on listed authors' designations.* designates facultative autogeny...... 53

APPENDIX TABLE 1. Descriptions of 2011 and 2012 sample sites. Definitions are based on OPIAM (Ontario Parks Inventory and Monitoring Program: Guidelines and Methodologies-Version 1.4 Draft, 25 May 2012) ...... 68

VII

LIST OF ABBREVIATIONS

The following abbreviations were used throughout the thesis:

Diptera: Culicidae

Ae= Aedes

Cq = Coquillettidia

An= Anopheles

Cs= Culiseta

Diptera: Tabanidae

H= Hybomitra

C= Chrysops

T=Tabanus

VIII

CHAPTER 1: GENERAL INTRODUCTION

The boreal forest covers a vast area in Ontario with few human settlements. Most of the region shows little of the ecological change and disturbance that accompanies human activity further south. Due to the largely untapped resources from this region, it is under increasing pressure for large-scale resource extraction. My study is part of a six year Far North Biodiversity Project (FNBP), a multispecies inventory conducted by the

Ontario Ministry of Natural Resources (OMNR). Parts of the study region are likely to be developed in the near future in the form of mining activity. The main goal of the FNBP study is to provide background data on species diversity for the far north region of

Ontario before development occurs. This will allow the effects of development to be quantified and will help inform land use planning and management strategies. My thesis provides accurate distributional information on two of the more important families of insects in this region, the Tabanidae (horse and deer flies) and Culicidae (mosquitoes).

Far North

The Far North of Ontario is “the area north of Woodland Caribou and Wabakimi

Provincial Parks, and north of the following forest management units: Red Lake Forest,

Trout Lake Forest, Lac Seul Forest, Caribou Forest, Ogoki Forest, Kenogami Forest, 1

Hearst Forest, Gordon Cosens Forest and Cochrane-Moose River” (Far North Act 2010).

The Far North of Ontario encompasses two Ecozones; the Ontario (Boreal)

Shield, and the Hudson Bay Lowlands. The Ontario Shield is a forested area with variable topography, drainage, geology and substrates dominated by near-surface bedrock. The Hudson Bay Lowlands Ecozone is a low elevation, relatively flat area with restricted drainage and therefore is mainly composed of peatlands and wetlands. The area

incorporates inland wetland habitats as well as the coasts of James Bay and Hudson Bay

(Crins et al. 2009).

There are few roads, with many only operational as winter roads, as the ground is too wet for the majority of the year to allow travel by vehicle, and all-weather road construction is difficult and expensive. As a result, this region is understudied. It is vast,

452,000km2, or 42% of Ontario’s land mass (Far North Science Advisory Panel 2010).

Habitats are mainly fens, bogs, or coniferous forests with a closed canopy (canopy cover greater than 60%). There is a great deal of standing water, although some areas are wetter than others (i.e., the lowlands). Snow and ice cover usually leave most of the far north of

Ontario, including my sample sites, in mid-May to early June and return again in mid-

October.

Biting flies in the far north

The combination of extensive wetlands and upland boreal forest provide ideal habitat for an immense number of two of the most important families of biting flies

(Wood 1985): horse and deer flies (Diptera: Tabanidae), and mosquitoes (Diptera:

Culicidae). Species from both of these families spread disease (Teskey 1970), and

torment humans and wildlife such as Moose (Alces alces), Woodland Caribou (Rangifer 2

tarandus), Black Bear (Ursus americanus), smaller mammals, birds, amphibians, and reptiles (Wood 1985). On the beneficial side, they also form an important part of the ecosystem because they are an abundant source of food for wildlife such as small mammals, birds, amphibians, and reptiles (Teskey 1970), and because they are nectar feeders and therefore could also act as secondary pollinators.

In northern Ontario, there are about 30 to 35 species in both families Tabanidae, and Culicidae, based on range maps in Wood et al. (1979), Teskey (1990), Darsie and

Ward (2005), Thomas and Marshall (2009), and Thomas (2011). The present diversity of wildlife in this region is a result of post-glacial colonization. Culicidae and Tabanidae are two families of biting flies with no non-native species present in this area (Wild Species

2010). Their abundance and wide distribution provides a good opportunity to understand general underlying patterns of species distributions that occur where there is limited human activity.

Thesis organization

This thesis has a general introduction, three research chapters, and a general discussion. Two of the research chapters summarize important range extensions and new range records that I found for both Culicidae (Chapter 2) and Tabanidae (Chapter 3) from an important remote region of northern Ontario that has been little studied and thus has few distributional records. In these chapters, I report numbers of each species I found (my fundamental data), the relative abundance of each species (a measure of the adaptations of each species to the site/environment) and diversity (a measure of the community of

insects based on two underlying measures, richness and abundance) of the western 3

(within 150 km of the First Nations communities near Big Trout Lake and Sandy Lake,

Ontario) and eastern study regions of the Far North Biodiversity Project (within 150 km of the First Nations community of Fort Albany, Ontario). To inform future biodiversity surveys, in Chapter 3 I also examine the trapping biases of passive Malaise traps and active sweep netting for collecting Tabanidae.

In Chapter 4, I use the trap bias information from Chapter 3 to examine several hypotheses about how the distribution patterns of Tabanidae might be affected by blood- feeding requirements. I discuss and summarize my work with a general discussion and conclusion.

To summarize, this research has three objectives:

1. To examine the species abundance, distributions and diversity in a little studied area with minimal human activity. This is important because it will act as a distributional baseline prior to any disturbance from industrial development. Baseline range information can be used to monitor the effects of climate change and development on the species, their distributions and the overall species composition within a changing region.

2. To examine a biogeographical hypothesis about post glacial colonization by insects in the boreal forest.

3. To examine hypotheses about anautogeny (i.e., blood-feeding) in Tabanidae, based on their spatial and temporal distribution, and trap biases.

CHAPTER 2: NEW RANGE RECORDS OF MOSQUITO SPECIES FROM

NORTHERN ONTARIO

(PUBLISHED IN 2013 IN THE JOURNAL OF THE ENTOMOLOGICAL SOCIETY OF ONTARIO 144: 3-14)

Authors: J.L. RINGROSE, K.F. ABRAHAM, D.V. BERESFORD

ABSTRACT

A survey for mosquitoes at 23 sites in the Ontario Shield and Hudson Bay

Lowlands of northern Ontario, Canada, in 2011 and 2012 yielded 19 species, including

16 of Aedes, and one each of Anopheles, Coquillettidia, and Culiseta. One species, Aedes pullatus (Coquillett) is newly recorded for Ontario. Eleven northern range extensions and one southern range extension are reported.

KEYWORDS: Culicidae, Northern Ontario, Hudson Bay Lowlands, species distribution

INTRODUCTION

The knowledge of distributions of many mosquito species (Diptera: Culicidae) in

Canada is incomplete. Jenkins and Knight (1952) conducted a survey of larval mosquitoes in southern James Bay. Steward and McWade (1960) published range summaries of species in Ontario. Wood et al. (1979) compiled the most complete account of mosquito distribution in Canada. The Canadian Endangered Species Conservation

Council (CESCC 2011) assessed the status of many species, including mosquitoes. Yet, areas such as northern Ontario are still relatively little sampled.

Northern Ontario has become the focus of increased mineral exploration and development (FNSAP 2010). Additionally, the area is projected to undergo significant ecological transformation over the next several decades due to climate change (FNASP

2010). Together, these two driving forces create a need for better knowledge of species’ distributions in northern Ontario before significant changes occur. A biological diversity survey of different taxa in northern Ontario was initiated in 2009 to address this issue

(OMNR 2012). The species composition and diversity information obtained will help determine land use, management and conservation planning, as well as provide baseline information to determine the impact of mining, forestry, and climate change.

Mosquito species lists for particular geographic areas include species that have not been collected there but are assumed to be present based on information from adjacent areas (e.g., Wood et al. 1979; Darsie and Ward 2005). Thus, it is reasonable to expect species to be found in northern Ontario if they have been found in similar habitat and at similar latitudes elsewhere, i.e., in spite of regional climatic differences, we expected to find species that have existing records from both adjacent western Quebec and northern Manitoba because of the large scale continuity of the ecosystems in the

boreal and subarctic forests that span these three provinces. For mosquito species whose 6

known distributional limits were either south or north of our study areas, we expected to extend known ranges north or south, respectively. Following this reasoning, and based on the range maps provided by Wood et al. (1979) and Darsie and Ward (2005), we predicted a maximum of 31 species in our surveys. In this paper, we report new information on occurrences of known species (range extensions), new collection locations and records of species new to the province for Culicidae from surveys of

previously unexplored areas of the far north of Ontario. We use both rarefaction and a lognormal analysis to explore the maximum number of species predicted in these areas and to gauge their relative abundances.

MATERIALS AND METHODS

Sampling took place within two different northern Ontario ecozones: the Ontario

Shield and Hudson Bay Lowlands ecozones (Crins et al. 2009) in 2011 and 2012, hereafter referred to as the western and eastern study areas, respectively, as part of a larger biological survey of animal and plant taxa undertaken by the OMNR (2012). The

2011 sampling occurred within 150 km of the First Nations communities near Big Trout

Lake and Sandy Lake, Ontario in the western study area. The 2012 sampling occurred within 150 km of the First Nations community of Fort Albany, Ontario, in the eastern study area. In each year, 12 sample sites were randomly selected from the computer generated grid of National Forest Inventory (NFI) points (Gillis et al. 2005). Actual sample locations sometimes differed by as much as 15 km from the NFI coordinates depending on feasibility of landing a helicopter. Our plot locations are the sites at which

field camps were established (OMNR 2012). Sample locations were within 1 km of the 7

field camp, which was verified using a handheld GPS (Garmin Rino 530HCx, NAD83, -

±3m accuracy).

Habitats at these sites were dominated by coniferous and shrub wetlands comprised largely of black spruce (Picea mariana (Britton, Sterns & Poggenb.)) and tamarack (Larix laricina (Du Roi) K. Koch) as well as shrub and sedge fens, and sphagnum bog. In 2012, Hudson Bay Lowlands sites had generally more standing water

than the western sample sites from 2011. Sampling occurred from 29 May to 17 July in

2011 and 4 June to 15 July in 2012.

In both years, the mosquito component of the sample regimen included daily sampling both by individual collection (ad hoc, when mosquitoes were present, approximately 30 minutes total), and a dusk and dawn sweeping with an insect net for 6 minutes at each sampling location. Individual collection consisted of catching mosquitoes that landed on the face, arms and legs of field crew members using snap cap vials (2.0 ml) before they had a chance to bite. These collections occurred throughout the day and late evening. Individual specimens in snap vials were preserved dry in the capture vials.

Adult mosquitoes collected by sweeping were placed in labeled sample jars with a silica desiccant to prevent deterioration from moisture. Many of the mosquitoes caught by sweeping had scales on their thoraces abraded and so could not be identified to species.

Therefore, more effort was placed on individual collection in 2012.

All specimens were pinned and identified by JLR and DVB using the keys of Wood et al.

(1979), and Thielman and Hunter (2007). Nomenclature was based on the WRBU Online

Catalog (2013). Voucher specimens were assigned individual specimen numbers (Table

2.2) and are stored at the Trent University Biology Department in Peterborough, Ontario. 8

Some vouchers are deposited in the Canadian National Collection of Insects, Ottawa.

Interpretations of new records and range extensions are based on comparison with range maps in Wood et al. (1979).

Analysis

Rarefaction analysis for the 2011 and 2012 catch data was performed using software on the University of Alberta website

(http://www.biology.ualberta.ca/jbrzusto/rarefact.php). This method relates sampling effort to number of species caught. The total number of species caught each year is used to calculate the expected number of species (with standard deviation) that would have been caught if fewer mosquitoes were sampled overall. Different species numbers for the same total catch sizes indicate community differences such as those due to site (e.g., habitat or phenological) or procedural differences.

We also fit the catch data (Table 2.1) to a lognormal distribution using the sum of squares method, i.e. Preston's method as described in Ludwig and Reynolds (1988). This allowed us to calculate the expected number of species by estimating the number of rare species not found in the samples. Essentially, it assumes that species of low abundance, e.g., about 1 per 1000 individuals, will only be found if at least 1000 individuals are collected. The lognormal distribution uses the abundance of different species and groups them into octaves or doubled catch classes, e.g., 0–1 individuals, 1–2 individuals, 2–4 individuals, 4–8 individuals and so on, and fits these frequencies to a lognormal curve by aligning the mode. Species that had only one individual caught could go into either the first or second class, so the number was divided between these classes, e.g., if one catches

5 species with only one individual each, then half of these (2.5) are assigned to the 0-1 9

class, and 2.5 to the 1-2 class (Ludwig and Reynolds 1988). One of the assumptions of this method is that very rare species will not be sampled, but can be calculated from the area of the normal curve to the left of the 0–1 class or veil line. The biological interpretation is that this class (0–1) would become the 1–2 class if our total catch size was increased. This analysis requires an iterative method to find values for two

parameters that provide the best fit: a (width), and So (height). We used the SOLVER optimization add-in function in Microsoft Excel 2007 version for this task.

RESULTS

We caught 896 mosquitoes in 2011 and 826 in 2012. Mosquitoes caught directly from the face and arms and housed in vials could all be identified to species, whereas only 117 (13%) of individuals from 2011 and 192 (21%) from 2012 sweeping could be identified to species. Species collected and collection locations are summarized in Tables

2.1 and 2.2. Twelve species were collected in the western study area in 2011 and 16 species in the eastern study area in 2012 (Table 2.1, Fig.2.1). The most abundant species identified in both years was Coquillettidia perturbans (Walker). Rare species i.e., those represented by a single individual collected in either year were Ae. cinereus Meigen, Ae. nigripes (Zetterstedt), Ae. provocans (Walker), Ae. pullatus (Coquillett), Ae. rempeli

Vockeroth and Culiseta impatiens (Walker).

Fitting to the lognormal distribution (Fig. 2.2), the expected number of species was 14.75 from the 2011 catches (fitted parameters a = 0.24, So = 2.0, Chi sq = 1.23, p = 0.94, d.f. =

5) and 23.4 species in the 2012 catches (fitted parameters a = 0.225, So = 2.97, Chi sq =

5.46, p = 0.36, d.f. = 5). By combining the totals of both years, our expected number of 10

species for northern Ontario was 28.2 species (fitted parameters a = 0.21, So = 3.35, Chi sq = 2.94, p = 0.82, d.f. = 6).

New Ontario record

Ae. pullatus has two distinct distributions, an eastern population in northern

Quebec and Labrador and a western population in Alberta, British Columbia, and the

Yukon (Wood et al. 1979). The single specimen we collected in the eastern study area is the first record in Ontario and extends the range of the eastern population westward.

Northward range extensions

Ae. canadensis (Theobald) is a widely distributed species found in forested regions of all Canadian provinces and the Yukon (Steward and McWade 1960). It is known to be found in Moosonee and Moose Factory in Ontario. Our collection was in the eastern study area.

Ae. cinereus is a common species in Ontario and has been found in Moosonee,

Moose Factory and the town of Kenora (Steward and McWade 1960). Jenkins and Knight

(1952) noted that Ae. cinereus was the most common larval species that they collected in the southern James Bay area but, oddly, they collected no adults. Our single specimen was collected in the eastern study area.

Ae. dorsalis (Meigen) is a rare northern species and in Ontario has only been collected in Moosonee and Moose Factory (Steward and McWade 1960). It was only collected in the eastern study area, which is relatively close to these communities.

Ae. implicatus (Vockeroth) is common in the northern and central parts of Ontario

and has been collected in Moose Factory (Steward and McWade 1960). It was collected 11

in both study areas.

Ae. excrucians (Walker) is found throughout North America (Wood et al. 1979).

It was collected by Jenkins and Knight (1952) in Moose Factory and Moosonee and by

Steward and McWade (1960). Our collection from the western study area provides a record for the gap between the eastern James Bay coast and Manitoba.

Ae. intrudens Dyar is found south of the tree line in late spring (Wood et al.

1979). It has been recorded from all provinces (Steward and McWade 1960). The species was common in the eastern study area, but was not found in the western study area.

Ae. provocans is a forest species and is a southern species in Ontario (Wood et al.

1979).The most northern record is from Great Slave Lake, Northwest Territories. We collected a single specimen in the eastern study area.

Ae. rempeli is one of the rarest of Canadian species (Vockeroth 1954). However,

Wood et al. (1979) suggested that this species may be widely but sparsely distributed in northern Ontario. We caught a single specimen along the Albany River about 150 km upstream from the James Bay coast.

Anopheles earlei Vargas is the most common species of this genus in Ontario.

Our collections of this species in both study areas extend the known range.

Cq. perturbans is common in southern Ontario (Wood et al. 1979). Jenkins et al.

(1952) found that this species was very abundant in a spruce forest west of Cochrane,

Ontario. In both study areas it was our most abundant species.

Cs. impatiens is a northern species usually found in forested regions and has been

recorded from Moose Factory (Steward and McWade 1960). Our single specimen came 12

from the western study area, providing a westward extension of the known range.

Southward range extensions

Ae. nigripes is an arctic species whose range, according to Wood et al. (1979), did not extend southward into Ontario. However, one recent record exists from Polar Bear

Provincial Park (Beresford 2011). One specimen was collected in the western study area in 2011, farther south than Polar Bear Provincial Park.

Range gap infills

Ae. abserratus (Felt and Young) is an uncommon species in Ontario (Wood et al.

1979). Steward and McWade (1960) reported the species from Moose Factory. Beresford

(2011) collected it in Polar Bear Provincial Park. Our collection of this species in both study areas fills the gap.

Ae. communis (De Geer) is one of the most widely distributed species in the northern hemisphere. Beckel (1954) stated that this species was rarely collected in the

Churchill area of Manitoba because it is non-biting in that area. In Ontario, records show it to be generally present and often abundant throughout the province. It has been previously found in Moose Factory (Steward and McWade 1960). This species was well represented (9.4%) in our collections from the western study area, but less so (1%) in the eastern study area.

Ae. hexodontus Dyar has been collected in Churchill, Manitoba both as larvae

(Vockeroth 1954) and as adults (Beckel 1954), and also from western Quebec and western Ontario (Wood et al. 1979). Our collection fills the gap.

Ae. impiger (Walker) is generally found in Nunavut and the Northwest Territories

(Steward and McWade 1960). It has been caught in Ontario at Moose Factory and along 13

the Albany River (Steward and McWade 1960) and in Manitoba at Churchill (Downes

1965). Our collections from our western study area fill a gap between Churchill and the

James Bay coast in Quebec. Surprisingly, we did not find any in our eastern study sampling sites, which are close to James Bay.

Ae. pionips Dyar is found in the forests of central and northern Canada, and has been collected from Moose Factory, Ontario (Steward and McWade 1960) and Churchill,

Manitoba (Beckel 1954). Not unexpectedly, our collections fill the gap.

Ae. punctor (Kirby) is a common species in Ontario and throughout Canada

(Steward and McWade 1960). Previous records are from Moosonee (Jenkins and Knight

1952) and Churchill, Manitoba (Beckel 1954). Our collections are within the expected range but fill distributional gaps in northwestern Ontario.

DISCUSSION

As expected we produced new distributional records, including both northward and southward range extensions, and filled gaps in former documented ranges. All of the species we collected are considered by CESCC (2011) to be secure (relatively widespread or abundant), except for five with undetermined status: Ae. impiger, Ae. implicatus, Ae. pionips, Ae. rempeli and An. earlei.

The rarefaction analysis, which standardizes across different sample sizes, indicates that the eastern region (2012) had slightly more species present than the western region

(2011). For example, in collections of 100 individuals we would only have been able to

catch about 13 species in the east compared to 11 in west (Fig. 2.1). The lognormal 14

analysis shows the same pattern, with 23.4 species predicted to be in the eastern region compared to 14.75 in the western region (Fig. 2.2). These analyses reveal that this difference in species richness may be a function of the different regions (e.g. habitats) rather than catch effort. The 2012 eastern Ontario collections were from sites with lower elevations (between 1 m to 88 m ASL) than the western 2011 sites (between 148 m and

379 m ASL). However, because these two regions were sampled in different years, we cannot attribute this difference to sampling region alone.

From our survey of the range maps we expected to find up to 31 species. Fitting the lognormal distribution to our overall catch numbers, our expected number of species was 28, a good estimate of species richness of this region.

In fact, we found only 19 species and four of the species we did catch were not expected from the range map analysis: Ae. nigripes, Ae. provocans, Ae. pullatus, Ae. rempeli. This means that 16 species from the range map analysis were expected but not found, either due to our sampling methods, phenology, or habitat preferences. Of these,

Wyeomyia smithii (Coquillett) is fully autogenous and has not been reported blood- feeding; Ae. diantaeus Howard, Dyar and Knab is not found in coniferous forests, Ae. spencerii (Theobald) is not found in forest regions; Ae. sticticus (Meigen) is generally restricted to floodwaters of rivers; Cs. morsitans (Theobald) and Culex restuans Theobald prefer to blood-feed from birds; Cx. territans Walker prefers reptiles and amphibians; Cs. alaskaensis (Ludlow) and Ae. mercurator Dyar are early spring species; An. walkeri

(Theobald), Ae. vexans (Meigen) and Ae. campestris Dyar & Knab are primarily

nocturnal biters. The remaining four of the expected species are rare, Ae. riparius Dyar & 15

Knab, Ae. flavescens (Müller), Ae. fitchii (Felt & Young) and Ae. decticus (Howard, Dyar

& Knab) (Wood et al. 1979).

All collection methods have inherent biases (Muirhead-Thomson 1991). Some important limitations to this survey are that collections occurred at randomly chosen sites

(i.e., not selected for high probability of detecting mosquitoes) and using simple methods that were part of a larger diversity survey. The mosquito portion of that survey was

limited by the logistics of available time and equipment at these remote sites. A collection effort that focused on targeting mosquitoes alone, within specific habitats, would likely have produced more of the expected species, and the use of CO2 traps or

CDC light traps would have produced far larger collections. Nevertheless, this study, despite its limitations, indicates that simple surveys undertaken in under-sampled regions can produce important baseline information that extends the previously known ranges.

18 2012 16 14 12 2011 10 8 2011 2012 6

100 200 300 4 km

expected number of species of expected number 2 0 0 50 100 150 200 catch size

FIGURE 2.1. Rarefaction analysis of mosquito catches (means and SDs) within 150 km 17

of Big Trout Lake and Sandy Lake in 2011 (closed circles) and within 150 km of Ft.

Albany in 2012 (open circles). The inset map of Ontario shows the sampling locations in

2011 and 2012.

3 2 2011 1 0 4 1 3 5 7 9 11 13 15 17 19 21 23 25 27

3 2012 2

number species/class of 1 0

1 2 4 8

- - - -

32 64

0 1 2 4 - - -

veil line 128

16 1 3 5 7 9 11 13 15 17 1932 21 23 25 27

64 catch class (octaves)

FIGURE 2.2. Fitted lognormal distributions of mosquito catches within 150 km of Big

Trout Lake and Sandy Lake in 2011 and within 150 km of Ft. Albany in 2012. The area of the region left of the veil line represents species that were too rare to be sampled with

18 our methodology.

TABLE 2.1. Culicidae species collected in 2011 within 150 km of Big Trout Lake and Sandy Lake, and in 2012 within 150 km of Fort

Albany.

Catch per year Date(s) captured Distribution change for Ontario

Species 2011 2012 2011 2012

Aedes abserratus (Felt and Young) 1 29 June 17–July 7 June 8–July 7 gap infill

Aedes canadensis (Theobald) 2 June 25, 28 new northern record

Aedes cinereus Meigen 1 June 23 new northern record

Aedes communis (De Geer) 11 2 June 10, 11 June 15, 28 northwestern gap infill

Aedes dorsalis (Meigen) 4 June 10 new northern record

Aedes excrucians (Walker) 2 2 July 12 June 23, 25 new northern record

Aedes hexodontus Dyar 4 11 June 2–July 12 June 8–13 gap infill

Aedes impiger (Walker) 18 June 2 northwestern gap infill

Aedes implicatus Vockeroth 2 3 June 2–17 June 8, July 10 new northern record

Aedes intrudens Dyar 19 June 8–26 new northern record

Aedes nigripes (Zetterstedt) 1 July 7 new southern record

Aedes pionips Dyar 28 32 June 2–July 7 June 8–July 14 gap infill

Aedes provocans (Walker) 1 June 8 new northern and eastern record

Aedes pullatus (Coquillett) 1 June 10 first record for province

Aedes punctor (Kirby) 7 35 June 6–July 3 June 8–July 7 gap infill

Aedes rempeli Vockeroth 1 June 26 new northern record for Ontario

Anopheles earlei Vargas 3 5 July 3 June 8, July 13 new northern record for Ontario

Coquillettidia perturbans (Walker) 39 44 July 3 to 15 June 17 to July 13 new northern record for Ontario

Culiseta impatiens (Walker) 1 June 6 new northern record for Ontario

TUIC # Year 2011 2012

15 15

- -

July 5 July 5 July 2 July 2

- - - -

June 7 June 7

June 23 June 23 July 21 June 18 June 18 June 25 June 25 July 9 July 9 July 16

June 16

- -

July 13 July 13

June 11 June 11

------

- -

June 8 June 8

uly 6 uly

July 3 July 3

J July 6

June 28 June 28 June 26 June 26

May 31 May 31

July 14 July 10

Sampling dates

June 5 June 5

July 10

June 16 June 16 June 12 June 12 June 19 June 19

(West)

82° 82° 8' 2"

89° 89° 6' 2 92° 1' 44" 93° 2' 33" 91° 9"49' 92° 4"46' 82° 2"41' 82° 2"49' 83° 2' 25" 82° 3' 24"

Longitude

89° 89° 43"40' 88° 51"54' 90° 38"21' 88° 33"33' 94° 38"13' 93° 10"32' 81° 48"57' 82° 13"39' 80° 11"23' 81° 57"50' 81° 23"39' 83° 24"17' 83° 44"22'

° 50"47'

(North)

Latitude

53° 53° 8"12' 54° 9' 30" 54° 1"27' 53° 9"36' 51° 8"39' 51° 8"58'

54° 54° 50"25' 53° 35"45' 54° 19"28' 53° 40"27' 52° 28"49' 52° 37"27' 53° 12"44' 52° 35"46' 51° 53"55' 51° 40"26' 52° 25"53' 51° 53"29' 52° 27"28' 52° 20"23' 51 52° 23"18' 51° 22"21' Species 0001 Aedes abserratus X X X X X X X X X 0002 Aedes canadensis X 0003 Aedes cinereus X 0004 Aedes communis X X X 0005 Aedes X

dorsalis 0006 Aedes excrucians X X X 0007 Aedes hexodontus X X X X X X 0008 Aedes impiger X 0009 Aedes implicatus X X X 0010 Aedes intrudens X X X X X 0011 Aedes nigripes X 0012 Aedes pionips X X X X X X X X X X X X X X X 0013 Aedes provocans X 0014 Aedes pullatus X 0015 Aedes punctor X X X X X X X X X X 0016 Aedes rempeli X 0017 Anopheles earlei X X X 0018 Coquillettidia perturbans X X X X X X X X 0019 Culiseta impatiens X

CHAPTER 3: NEW RANGE RECORDS, AND A COMPARISON OF SWEEP NETTING AND MALAISE TRAP CATCHES OF HORSE FLIES AND DEER FLIES (DIPTERA: TABANIDAE) IN NORTHERN ONTARIO (PUBLISHED IN 2014 IN THE JOURNAL OF THE ENTOMOLOGICAL SOCIETY OF ONTARIO (Vol 145)

Authors: J.L. RINGROSE, K.F. ABRAHAM, D.V. BERESFORD

ABSTRACT

A survey of horse flies and deer flies (Diptera: Tabanidae) was conducted in northern

Ontario, Canada in 2011, at 11 sites, and 2012, at 12 sites using Malaise traps and daily sweep netting. We caught 2168 tabanids of 30 species from 3 genera: 10 Chrysops, 18

Hybomitra, and 2 Tabanus. Malaise traps caught fewer individuals than sweep netting but more species: 850 tabanids from 28 species, 8 of which were not caught sweep netting.

Sweep netting caught 1318 tabanids from 22 species, with 2 not found in Malaise trap samples. Studies of insect diversity should incorporate a variety of sampling methods.

We present the first record of Hybomitra osburni (Hine) in Ontario. In addition, we also present 10 northern range extensions and 3 western range extensions.

INTRODUCTION

When habitats change, insect populations respond rapidly, increasing or decreasing, depending on the species characteristics. These changes occur across a range of temporal and spatial scales, and are unique for each species. The ability to increase their populations rapidly makes insect diversity an efficient indicator for monitoring both short and long term environmental changes (Danks 1992). Insects are known to track environmental change (Niemela et al. 1993). The variety of habitats occupied by the vast

number of insect species in Ontario means that studies that require tracking environmental change can benefit from using some insect group for monitoring change.

For such work, up to date distributional data are needed for a variety of insect species.

The "Ring of Fire" region in northern Ontario contains large deposits of chromium and other minerals (Far North Science Advisory Panel 2010), and anticipated large-scale extraction processes will likely alter insect diversity and distribution at some scales. To see the effect of such development, as well as possible effects of changing climates, baseline entomological records in these areas are needed. With this purpose, the

Ontario Ministry of Natural Resources (OMNR) started a project in 2009 designed to survey insect diversity and establish species distributions in Northern Ontario.

A widespread and easy to find group is Tabanidae which includes horse flies and deer flies. The first comprehensive report on the Tabanidae of Ontario was compiled by

Pechuman et al. (1961). This was followed by a more complete treatment of Tabanidae in

Canada by Teskey (1990). Since then, two pictorial keys, one on deer flies (Thomas and

Marshall 2009) and one on horse flies (Thomas 2011), added new range information for this group. Further reports of sampling continue to add distributional data to our

knowledge of Tabanidae (e.g., Beresford 2011). 24

In this paper, we report on range extensions of several species of Tabanidae in northern Ontario collected in 2011 and 2012. Our surveys were conducted using two sampling methods, sweep netting and Malaise trapping. We report on the different species caught by these two sampling methods, as well as new range information for several species.

MATERIALS AND METHODS

We sampled horse flies and deer flies in northwestern Ontario in 2011 and northeastern Ontario in 2012 (Fig. 3.1, inset map) at 12 locations each year. Site locations are provided by Ringrose et al. (2013). Generally, sampling took place within 1km of remote field sites that were accessed by helicopter. The 2011 sampling was completed in the western half of the Ontario boreal forest about 150 km of the First Nation communities of Big Trout Lake and Sandy Lake. The 2012 sampling occurred in the eastern part of the Ontario boreal forest bordered by James Bay within 150 km of the

First Nations community of Fort Albany, Ontario. Sampling dates were 5 June to 17 July

2011 and 5 June to 15 July 2012.

Tabanids were sampled daily by two methods, Malaise traps (6m trap model no.

2877, BioQuip Products 2321 Gladwick Street, Rancho Dominguez, CA 90220, USA), and sweep netting. The collecting heads of the Malaise traps were filled with 80% denatured ethanol to kill and preserve caught tabanids. These were emptied and replaced daily at 21:00.

Sweep netting was completed at midday as the surveyor (JLR) walked slowly,

sweeping for 5 minutes through any Tabanidae that assembled around the researcher. The 25

netted tabanids were killed by placing the end of the net in large killing bottles charged with acetone. Specimens were then removed from the net and stored in bottles filled with

80% denatured ethanol. The ethanol in each storage bottle was replaced once after 24 hours.

All tabanids were pinned and identified by JLR and DVB using the keys found in

Teskey (1990), Thomas and Marshall (2009) and Thomas (2011). The main pinned

collection is stored in insect cabinets at Trent University, Biology Department,

Peterborough, Ontario. A reference collection of voucher specimens is housed at the

Canadian National Collection of Insects, Arachnids and Nematodes, Ottawa.

Analysis

A list of the expected species was produced from the distribution records reported in the keys listed above. For those species that did not have records in northern Ontario, we reasoned that any species with records that straddled northern Ontario (either east and west, or north and south of the sampling regions) was likely present in northern Ontario.

We compared this expected number of species to the predicted number which we calculated using the lognormal distribution method (Preston's method as described in

Ludwig and Reynolds 1988). This approach allows one to predict the number of species present in an area using sampling data. It is based on a general observation that most species are more or less moderately abundant (the middle region of the lognormal distribution), a few species are very abundant (forming the right tail of the lognormal distribution) and a few are very rare (the left tail of the lognormal distribution). In practice, it enables one to predict the number of rare species that are expected but which

might be missed, depending on sampling intensity. In scenarios where there is limited 26

range data there can be a difference in expected numbers between the lognormal method and range map analysis. Parameters for the lognormal model were fitted using the

SOLVER function in MICROSOFT EXCEL 2007.

Catch data were analyzed using the online rarefaction calculator from the

University of Alberta (http://www.biology.ualberta.ca/jbrzusto/rarefact.php), to

determine the effects of collection size on the number of species collected, as well as to compare trapping methods.

RESULTS

From our assessment of the published range maps, we expected 31 species of

Tabanidae to potentially be present: 23 previously known from northern Ontario in the regions where we conducted our study (11 Hybomitra, 8 Chrysops, 2 Atylotus, 2

Tabanus), and 8 species with ranges that straddle our study regions (5 Hybomitra, 2

Atylotus, and 1 Haematopota). In our survey we found 30 species over the two years: 18

Hybomitra, 10 Chrysops, and 2 Tabanus, but no Atylotus or Haematopota. We collected

2168 tabanids: 839 from 24 species in northwest Ontario (2011 sampling), and 1329 from

25 species in northeast Ontario (2012) (Fig. 3.1. Tables 3.1 and 3.2).

The expected number of species was calculated to be 26 (lognormal fitted parameters, α = 0.28, So = 4.09, χ2 = 7.93, p = 0.34, d.f.=7) in the northwestern collections (2011) and 28 (fitted parameters, α = 0.24, So = 3.73, χ2 = 9.89, p = 0.27, d.f.

= 8) in the northeastern collections (2012), and 33 species for the combined data set

(fitted parameters, α = 0.23, So = 4.27, χ2 = 7.72, p = 0.56, d.f. = 9).

The three most abundant species caught in the northwest (2011) were Chrysops 27

excitans Walker (35%), Hybomitra epistates Osten Sacken (21%) and Hybomitra lurida

(Fallén) (19%). In the northeast (2012) the most abundant were Hybomitra affinis (Kirby)

(33%), Chrysops. excitans (22%) and Hybomitra lurida (19%).

New Ontario record

Our collection of 3 individuals of Hybomitra osburni (Hine) (2 in 2011 and 1 in

2012) is, to our knowledge, the first records of this species in Ontario. This species has

been collected in all western provinces and the Yukon territories (Teskey 1990) but was previously not known to occur east of Manitoba.

Range extensions

Our study provides new northern range records in Ontario for three species of

Chrysops: C.cuclux Whitney, C. niger Macquart, and C. venus Philip; six species of

Hybomitra: H.criddeli (Brooks), H. epistates, H. lasiophthalma (Macquart), H. pechumani Teskey & Thomas, H. tetrica (Marten), H. trepida (McDunnough); and one species of Tabanus, T. vivax Osten Sacken. In addition, we provide three new western records of Hybomitra for Ontario: H. minuscula (Hine), H. typhus (Whitney), and H. frosti Pechuman.

Gap infills

Chrysops ater Macquart is described as an abundant species having a general northern distribution in Canada south of the tree line (Teskey 1990). Our collection of a single specimen in 2012 is therefore not a surprise, however, there has been little collection in northern Ontario so our collection has filled a gap between previous collecting locations. It is perhaps surprising that it is so rare in our collections. Our

records of C. dawsoni Philip and C. frigidus Osten Sacken are consistent with known 28

ranges.

C. excitans, and C. mitis Osten Sacken, and to a lesser extent C. nigripes

Zetterstedt and C. zinzalus Philip, are found in Canada south of the tree line (Teskey

1990), and have recently been reported from Polar Bear Provincial Park (Beresford

2011). Our records are consistent with these reports.

We caught eight species of Hybomitra, consistent with known ranges: H. affinis, the most abundant and widely distributed Canadian species of Tabanidae (Teskey 1990;

Thomas 2011), H. arpadi (Szilady), H. frontalis (Walker), H. hearlei (Philip), H. illota

(Osten Sacken), H. lurida, H. nuda (McDunnough), and H. zonalis (Kirby).

Tabanus marginalis Fabricius has been collected from across Canada except on the east coast (Teskey 1990). While the known range encompasses our sampling locations (northern Manitoba and northern Quebec) our records are the northernmost from Ontario.

Trap comparison

The Malaise sampling caught fewer individuals than sweeping yet produced more species than sweep netting: Malaise traps, 850 individuals from 28 species (151 in 2011 and 699 in 2012); sweep netting, 1318 individuals from 22 species (688 in 2011 and 630 in 2012) (Fig. 3.2). Two species, C. cuclux and H. nuda, were absent from Malaise traps, and eight species, C. ater, C. frigidus, C. niger, C. venus, C. zinzalus, H. frosti, H. hearlei, and T. vivax were absent from the sweeps. When differences are examined within each year, the effect of trapping method becomes even more pronounced: 9 species

caught only by sweep netting and 5 only in Malaise traps in 2011; 1 species caught only 29

by sweep netting and 10 only in Malaise traps in 2012 (Table 3.1, Fig 3.2). H. affinis accounted for 16% of all tabanids caught in both years (354/2168, Table 3.1), yet 93% of all H. affinis were caught by netting, the remainder caught in Malaise traps. Similarly, C. mitis made up only 3% of all tabanids caught (72/2168), with 76% of C. mitis caught by netting, and only 24% caught in Malaise traps. Among the less common species, this

pattern was observed in H. lasiophthalma, C. dawsoni, H. frontalis, and H. pechumani

(Table 3.1).

DISCUSSION

All new range information and distributional locations are based on comparison with range maps from Teskey (1990), Thomas and Marshall (2009) and Thomas (2011) and is to the best of the authors’ knowledge at the time of publication. Because of the few intensive studies from this region we expected to add range records for many of the species we collected.

Our expected number of species from range map assessment (26) underestimated the number of species we caught (30). We collected 10 Chrysops versus 8 expected, 18

Hybomitra versus 16 expected and 2 Tabanus versus 2 expected). The lognormal projection of 33 species was three higher than what we found. Our survey did not include catches from August, and we expect there are more species in our study region that might emerge later in the season.

Perhaps the most interesting aspect of this study is the different species we caught

using the two trapping methods (Table 3.1). Some species were abundant in both trapping 30

methods, such as C. excitans, H. epistates, and H. lurida. In contrast, H. zonalis was abundant in the Malaise collections, while H. affinis and C. mitis were abundant in the netted samples. Insect survey methods are known to have inherent trapping biases

(Mihok 2002). It is possible that the difference between sweep and malaise catches are a result of yearly or locational variation. If there is an interaction between year or location and trapping method, the associated trapping biases would be more pronounced.

Therefore, these results highlight the importance of collecting using a variety of methods in insect surveys to overcome these catch biases.

Generally, the most abundant species were those that were caught over the longest period. There were exceptions to this pattern. In 2011, H. affinis was the most abundant species (275), but was only caught during 5 sampling sessions (Table 3.2) whereas H. arpadi (33) and H. lasiophthalma (21), both relatively uncommon, were also captured during 5 of the sampling sessions. In 2012, H affinis (79) was caught over 9 sessions, but was less abundant than H. lurida (252), which was caught over 7 sessions (Tables 3.1 and

3.2).

30 2011 25 2012 20

2011 10 2012

100 200 300 km

xpected number of of species number xpected 5 e

0 0 500 1000 1500 catch size

Malaise 2012 20 Malaise 2011

netted 2011 15 netted 2012

10 expected number of species 5

0 0 200 400 600 800 catch size

FIGURE 3.2. Rarefaction analysis of 2011 and 2012 Tabanidae collection data,

separated by trapping method. Malaise traps (circles) and sweep netting (diamonds) in

2011 (closed) and 2012 (open). Error bars represent standard deviations.

TABLE 3.1. Species and number of Tabanidae collected in 2011 and 2012 using Malaise traps and sweep netting, with abundance records.

Species 2011 2012 Total Malaise netted Malaise netted Chrysops ater Macquart 1 1 Chrysops cuclux Whitney 1 1 Chrysops dawsoni Philip 2 4 10 16 Chrysops excitans Walker 37 151 245 225 658 Chrysops frigidus Osten Sacken 5 1 6 Chrysops mitis Osten Sacken 11 50 6 5 72 Chrysops niger Macquart 1 1 Chrysops nigripes Zetterstedt 1 1 1 3 Chrysops venus Philip 1 1 Chrysops zinzalus Philip 4 8 12 Hybomitra affinis (Kirby) 7 268 17 62 354 Hybomitra arpadi (Szilady) 8 25 27 25 85 Hybomitra criddlei (Brooks) 1 1 2 Hybomitra epistates Osten Sacken 14 124 153 291 Hybomitra frontalis (Walker) 5 9 14 Hybomitra frosti Pechuman 2 2 Hybomitra hearlei (Philip) 2 2 Hybomitra illota (Osten Sacken) 3 2 1 6 Hybomitra lasiophthalma (Macquart) 21 4 2 27 Hybomitra lurida (Fallén) 59 104 131 121 415 Hybomitra minuscula (Hine) 6 9 3 2 20

Hybomitra nuda (McDunnough) 14 14 Hybomitra osburni (Hine) 2 1 3 Hybomitra pechumani Teskey & Thomas 4 1 8 13 Hybomitra tetrica (Marten) 2 1 3 Hybomitra trepida (McDunnough) 13 19 7 39 Hybomitra typhus (Whitney) 3 8 11 Hybomitra zonalis (Kirby) 2 5 67 6 80 Tabanus marginalis Fabricius 8 1 9 Tabanus vivax Osten Sacken 7 7 Total 151 688 699 630 2168 no. species 15 19 24 15 30

TABLE 3.2. Tabanidae species listed for each sampling location and date. Only 11 sampling sites are included for 2011 as one additional planned sampling site was unable to be surveyed due to weather.

Year 2011 2012

June 16 June 16 June 12 June 12 June 19 June 19 June

July 1 July

June 5 June 5 June

Sampling dates Sampling

July 14 July 10 July

May 31 May 31 May

June 28 June 28 June 26 June 26 June

July 6 July 6 July

July 3 July 3 July

June 8 June 8 June

- -

------

June 11 June 11 June

July 13 July 13 July

- -

June 16 June

Ju 23 June 21 July 18 June 18 June 25 June 25 June 9 July 9 July 16 July

June 7 June 7 June

- - - -

July 5 July 5 July 2 July 2 July

- -

15 15

ne 23 ne

Longitude (West) Longitude

89° 40'89° 42" 54'88° 51" 21'90° 37" 33'88° 33" 13'94° 38" 57'81° 47" 39'82° 13" 23'80° 10" 50'81° 56" 39'81° 22" 17'83° 23" 40'89° 42" 54'88° 51"

89° 6' 27" 89° 1' 43" 92° 2' 32" 93° 49'91° 8" 32'93° 9" 46'92° 3" 41'82° 2" 82° 6' 27" 89° 1' 43" 92°

49' 1"

Latitude (North) Latitude

54° 25'54° 49" 45'53° 34" 28'54° 18" 27'53° 39" 49'52° 27" 27'52° 37" 44'53° 12" 46'52° 34" 55'51° 53" 26'51° 40" 53'52° 25" 29'51° 53" 28'52° 26" 25'54° 49" 45'53° 34"

53° 12'53° 8" 9' 29" 54° 27'54° 1" 36'53° 8" 39'51° 7" 58'51° 8 12'53° 8" 9' 29" 54°

Species C. ater X C. cuclux X C. dawsoni X X X X C. excitans X X X X X X X X X X X X X X X X X C. frigidus X X

C. mitis X X X X X X X X X X C. niger X C. nigripes X X C. venus X C. zinzalus X X X X H. affinis X X X X X X X X X X X X X X H. arpadi X X X X X X X X X X X X H. criddlei X X H. epistates X X X X X X X X X X X X H. frontalis X X X H. frosti X H. hearlei X X H. illota X X X X H. lasiophthalma X X X X X X X X H. lurida X X X X X X X X X X X X X X H. minuscula X X X X X H. nuda X X X H. osburni X X H. pechumani X X X X H. tetrica X X X H. trepida X X X X X X X H. typhus X X X H. zonalis X X X X X X X X X X T. marginalis X X X T. vivax X X X X

CHAPTER 4: ANAUTOGENY IN TABANIDAE

ABSTRACT

The differences in host-seeking patterns, from 2011 in northwestern and 2012 in northeastern Ontario, of autogenous and anautogenous Tabanidae species were examined.

We analyzed three hypotheses based on adult longevity, larval development and presence of anautogenous and autogenous species. In addition, we compared Malaise and netted samples in an attempt to determine trapping biases of anautogenous and autogenous species. Anautogenous species were present longer (22.5 versus 12.7days) and showed up earlier in collections than autogenous species both years (14 June versus 23 June and 16

June versus 27 June, in 2011 and 2012, respectively). Anautogenous and autogenous species had clumped temporal distributions (VMR= 14.73 versus 6.76 and 15.4 versus

5.6, in 2011 and 2012, respectively). However, there was no significant difference in the level of clumping between autogenous and anautogenous species. Anautogenous species were more prevalent in sweep netting catches and autogenous species were significantly more prevalent in the Malaise traps in 2011.The same trend was prevalent in 2012, but was not significant.

INTRODUCTION

Hematophagous (blood-feeding) Diptera including the biting flies, such as mosquitoes and horse flies, target reptiles and amphibians, birds, small and large mammals (including humans) for their sustenance (Turell et al. 2001; Teskey 1990).

While generally perceived negatively because they can transmit disease to wildlife in

Ontario's boreal forests (Hongoh et al. 2012), these insects are also an important source of food for a variety of birds, and insects such as dragonflies (Odonata) and many species of wasps (Hymenoptera, Teskey 1990). For the biting flies, obtaining a blood meal is thought to be an important part of a life history strategy to overcome a combination of habitat constraints such as short breeding season and nutrient-poor larval habitats, as well as providing a high quality source of protein for augmenting egg production, for example, by increasing the number of ovipositions (Thomas 1971). Various species have adopted strategies to deal with environmental and life history stresses, ranging from general to highly specialized host species preferences, to blood meal requirements for egg maturation or simply as needed for extra energy (Downes 1958).

The terms autogeny and anautogeny are used to describe the blood meal requirements of biting flies for viable egg production. Spielman (1971) defined autogeny in mosquitoes as “a situation where there is no specific food-mediated ovarian diapause” and defined anautogeny as “the necessity of a blood meal for development of eggs”.

Thomas (1971) defined autogeny as “a development of follicles beyond the diapausing stage without the female taking a blood meal”.

Autogeny can be facultative or obligatory. Facultative autogeny occurs when an 39

individual does not require a blood meal for egg production but will take a blood meal if one is available, whereas obligatory autogeny occurs when an individual cannot take a blood meal (Clements 1992). Autogenous species rely on lipid content obtained during the larval stages to lay their eggs (Teskey 1990). However, due to a lack of research there is some variation regarding the classification of species these two categories. There is

some uncertainty as to the classification of species in areas other than those of the authors’ study and many discrepancies between authors in terms of their classifications.

For this paper, we classify Tabanidae into those species that have been previously identified as requiring a blood meal before oviposition as anautogenous, and those species that do not require a blood meal as autogenous even if they will sometimes take blood (i.e., facultative autogeny). Classifications were based on Leprince and Maire

(1990), McElligott and Lewis (1998) and Thomas (1971, 1972) (Table 4.1).

Blood feeding is a risky undertaking for an insect due to host defensive behaviour.

For anautogeny (must have blood) to persist, there must be a benefit to the survival of the adult or the offspring that outweighs the associated risks. The most complete study on autogeny in tabanids was conducted by Thomas (1971), although others have also contributed (McElligott and Lewis 1998; Leprince and Maire 1990; Thomas 1972).

Thomas (1971) proposed several hypotheses to explain the existence of anautogeny among the Tabanidae.

In this paper, I test three distinct, but not necessarily mutually exclusive, hypotheses based on the arguments made by Thomas (1971) about why both autogeny

and anautogeny persist in Tabanidae. I also test a fourth hypothesis that explains 40

differences seen in host-seeking patterns of autogenous and anautogenous species. My test metric (trap catches using various trapping methods) is based on the argument that anautogenous species will need to be more persistent in obtaining a blood meal than autogenous species, and that this will be manifest in the trapping biases of each species.

Simply stated, anautogenous species will likely be found flying where blood meals can be obtained.

The hypotheses I tested are:

H1 Adults of anautogenous species live longer than autogenous species;

H2 Larvae of anautogenous species should have shorter development periods and can experience reduced larval predation;

H3 Anautogenous larvae can live in poor-nutrient habitat because they can compensate as adults;

H4 Anautogenous tabanids are more persistent host seekers than (facultative) autogenous species.

Rationale and consequences for support of above hypotheses

Rationale for H1 Adults of Anautogenous species live longer than autogenous species. The survival of autogenous and anautogenous species at the uniparous stage

(having laid one batch of eggs) ranged between 0 to 16.7% and 0 to 73.3%, respectively

(Thomas 1971) which suggests greater survival for anautogenous species. It is likely that autogenous species lack the required energy reserves required to complete more than one gonotrophic cycle (Thomas 1971). Since egg development is more or less at the same rate in the various Tabanidae species (4-6 days) (Teskey 1990), if anautogeny is an adaptive

trait based on increasing the number of ovipositions, this should be manifested through 41

increased adult longevity and ultimately longer capture periods. Possibly the best method to assess adult longevity would be a live mark-recapture of individuals throughout the season. Since this would be logistically impossible I used capture period as a proxy. We should be able to catch anautogenous species over a longer period of time than autogenous species from the same region. This pattern should hold regardless of habitat type and collection location.

Rationale for H2. Larvae of Anautogenous species should have shorter development periods and can experience reduced larval predation. A shorter larval period means fewer days exposed to larval predators. If predation rates and other sources of mortality are similar among species, independent of blood feeding category, this should mean higher larval survival. The trade-off is that anautogenous species will not be able to obtain the required nutrition for egg production without taking a blood meal as adults (Thomas 1971). Tabanid larvae are eaten by crane-fly larvae (Diptera: Tipulidae) and other Tabanidae larvae (Miller 1951). Shorter development times should be apparent in terms of the adult emergence dates, with anautogenous species adults generally being present earlier in the spring/summer collections than autogenous species. I test this hypothesis by examining emergence times.

Rationale for H3. Anautogenous larvae can live in poor-nutrient habitat because they can compensate as adults. The expression of anautogeny and autogeny may be directly linked to the nutritional quality of the larval habitat. Anautogenous species should be able to take advantage of low nutrient larval habitats as well as nutritionally rich larval habitats, whereas autogenous species would require high quality

larval habitat. In terms of abundance, if we assume that there is more good-plus-poor 42

quality habitat (i.e. both) than good habitat, we should catch higher numbers of adults from each anautogenous species than from autogenous species, reflecting higher, more widespread populations. The reason for this is that anautogenous species can use good and poor quality habitat and autogenous species can use only good quality habitat.

However, this could also be the hypothesis for facultative autogeny at the species level, which raises the interesting possibility that there is a continuum of autogeny (i.e.,

obligatory autogeny to facultative autogeny to anautogeny) with the middle type being an ecological equivalent to the obligatory (evolved) anautogenous type. If this is the case, then both autogenous and anautogenous species will be able to take advantage of both good and poor habitat, and we will not be able to find evidence for or against this hypothesis. This would be manifest by similar abundance and distribution of both autogenous and anautogenous species in my collections, in the absence of evidence for or against the other hypotheses.

Rationale for H4. Anautogenous tabanids are more persistent host seekers than (facultative) autogenous species. Host seeking has disadvantages such as increased predation and energy expenditure; therefore, horse flies that do not require blood meals

(autogenous) need not be as persistent in host seeking as species that need to blood-feed.

Malaise traps are designed to act as flight interceptors, whereas netted samples catch tabanids that come to the operator as a potential host. Using these two sampling methods, sweep netting should collect more proportionately anautogenous tabanids (individuals and species) than autogenous species compared to Malaise catches.

Predictions 43

Some of the predictions arising from these hypotheses are tested by sampling using both Malaise traps and sweep netting, and are measured when anautogenous and autogenous species are first collected, persistence in traps and their abundance throughout the sampling season.

P1 Anautogenous species will be collected over a longer period of time than autogenous species because in order to have more gonotrophic cycles they would need to be alive longer.

P2 Adults of anautogenous species will be caught earlier in the sampling season than autogenous species because anautogenous species can supplement nutrition as an adult by blood-feeding.

P3 Assuming low quality or marginal larval habitat is more common in the boreal region than high quality larval habitat, anautogenous species will be more widespread or even in catches since they can use high and low quality habitat, whereas autogenous species will tend to be clumped where larval habitat is the best because they can only use high quality habitat.

P4 Anautogenous species will be more abundant in netted sweep samples than in

Malaise trap samples, due to their attraction to the researchers.

MATERIALS AND METHODS

Sampling

Sampling occurred in northern Ontario in the same locations listed in chapters 2 and 3. Sampling was conducted from 5 June to 17 July in 2011 and from 4 June to 15

July in 2012 in northern Ontario. A single Malaise trap as well as handheld net sweeping was used at each location. Sweep netting was conducted at random intervals throughout the day (between 09:00 and 21:00), generally within the vicinity of camp (within 100m), and was conducted for a total of 5 minutes total per sampling event. Sweeping followed a figure 8 pattern through the main body of the insect cloud that assembled near the net operator.

Species Classification

We classified species as autogenous and anautogenous based on the reports of previous authors (Table 4.1). Species for which no data exist were excluded from these analyses.

Analysis

We determined the duration each species was present in collections using the date

of first and last appearance of the species each year. We used ANOVA to compare 45

species' durations (H1) and species' first collection dates (H2), using autogeny and year as factors.

To test temporal clumping for each species, we used each day's catch (combined sweep and Malaise trap) as my sample unit, for each species. We assessed aggregation patterns (clumping) using variance-mean ratios (VMR), and compared the mean autogenous species and anautogenous species VMRs using a t-test, assuming unequal

variances. The VMR is used to test for random, regular or clumped distributions. It is a characteristic of the Poisson distribution, used for modeling random events, that the variance equals the mean, so that the VMR = 1 in a Poisson distribution. In clumped distributions the variance is greater than the mean, so the VMR >1, and in regular or evenly spaced distributions, the VMR < 1 (Krebs 1989).

Finally, we compared the catches of anautogenous and autogenous species in both sweep netting and Malaise traps. For each year, we plotted the sweep netting catches (x- axis, ln + 1) against the Malaise trap catches (y-axis, ln + 1) for each species. We then used a multiple regression with a dummy variable to test whether the regression line of anautogenous species was lower (lower y-intercept) than the regression line of autogenous species, meaning that the anautogenous species tended to have higher sweep/Malaise catch ratios than autogenous species. This was done for each year's data.

We used Microsoft Excel 2007 and Statistica version 7 for ANOVA.

RESULTS

The mean (SD, n) durations for anautogenous and autogenous species presence over a sampling period were 22.3 (4.34, n = 9 species) and 15.4 (4.6, n = 8 species) days

in 2011, and 22.6 (4.9, n = 7 species) and 10.1 (4.1, n = 10 species) in 2012, respectively. 46

Generally, anautogenous species were present longer (mean = 22.5, SD = 3.29, days) than autogenous species (mean = 12.7, SD = 3.09, days) (Autogeny type F(1, 30) = 4.64, p

= 0.039). There was no effect of year nor was there any interaction between year and autogeny (Year: F(1, 30) = 0.31, p = 0.58, Interaction: F(1, 30) = 0.37, p = 0.54).

When we tested the date of first catch we found that anautogenous species showed up in collections earlier than autogenous species both years: Julian day 166.6 (14

June) (4.15 SD, n = 9 species) and Julian day 175.0 (23 June) (4.40 SD, n = 8 species), respectively, in 2011, and Julian day 168.4 (16 June) (4.7 SD, n = 7 species) and Julian day 179.1 (27 June) (3.93 SD, n = 10 species) respectively, in 2012. Generally, anautogenous species were caught earlier (Julian day 167.5, or 15 June) than autogenous species (Julian day 177.1 or 25 June) (F(1, 30) = 4.92, p = 0.034). Again, there was no effect of year nor was there any interaction between year and autogeny (Year: F(1, 30) =

0.48, p = 0.49, Interaction: F(1, 30) = 0.07, p = 0.80).

The average VMRs of collection date for anautogenous and autogenous species were 14.73 (SD = 31.2, n = 9) and 6.76 (SD = 11.9, n = 8) in 2011, and 15.4 (SD = 16.4, n = 7) and 5.6 (SD = 10.7, n = 10) in 2012, indicating that both anautogenous and autogenous species were clumped. There was no difference in the level of clumping between autogenous and anautogenous species (2011, t = 0.71, d.f. = 10, p = 0.25; 2012, t

= 1.39, d.f. = 10, p = 0.097; t-test unequal variances (Ruxton 2006)).

From the dummy regression analysis, anautogenous species were more prevalent in sweep netting catches (the y-intercept was lower, Fig 4.1), and autogenous species were more prevalent in the Malaise traps in 2011 (t-test of y-intercept heights, t = 2.67, p

= 0.018, n = 17, 9 anautogenous and 8 autogenous species) (Fig. 4.1). While the same 47

general trend was seen in 2012, it was not significant (t-test of y-intercept heights, t =

0.81, p = 0.43, n = 18, 10 anautogenous and 7 autogenous species) (Fig. 4.2).

DISCUSSION

Evidence was consistent with prediction 1, that anautogenous species would be collected over a longer period, which supports the first hypothesis, and might provide evidence that anautogeny might allow a higher reproductive output due to enabling more

than one gonotrophic cycle. However, further analysis using gonotrophic cycles directly is needed. This trend towards longer periods of collection occurred in both regions (west boreal 2011 and east boreal 2012). In my study, sampling was not conducted past July, so it is possible that autogenous and/or anautogenous species might have continued past the collection date. However, Tabanidae numbers in the Ontario boreal forest further south of my study sites peak in early July, and August collections tend to be low (Raponi 2014, unpublished thesis). My work spanned the major period when adult Tabanidae were present. One could argue that anautogenous species have multiple generations each year.

However, Teskey (1990) suggests that larval development takes 9-10 months suggesting only one generation a year is possible.

The second prediction was also supported, which is evidence for the shorter larval period hypothesis (H2). However, larval period was not analyzed directly. Anautogenous species were collected significantly earlier than autogenous species. Indeed, the earlier catches of the anautogenous species is the reason that anautogenous species were in collections over a longer period of time. Temperature cues are known to be drivers of insect development (Beresford and Sutcliffe 2008) and I made the assumption that

anautogenous and autogenous species started yearly development at the same 48

temperature/time.

Prediction 3 was not supported with my data, providing no support for the habitat hypothesis. I did not find differences in clumping between the two feeding types of flies, suggesting that both forms are equally clumped spatially across the study areas. Because of these results, the hypothesis that I think is best supported by the data is that

anautogeny persists in Tabanidae in northern Ontario species to reduce the larval development period.

Why? The most likely reason is to reduce predation at this vulnerable stage.

Tabanidae larvae are preyed upon by birds, rodents, fish and many terrestrial and aquatic insects (Burger 1977). Miller (1951) found that the smaller Chrysops larval mortality was

30%, mostly from predation by crane flies and larger Tabanus larvae. It is likely that cannibalism (within a species) and predation (within the Tabanidae family) are the major predation risks of immature tabanid larvae (Burger 1977). Larval tabanids have also been found with nematode parasite infestations (Burger 1977).

It is possible that the expression of autogeny acts on a scale where in high quality habitat they are obligatory autogenous, in marginal or poor habitats they are facultatively autogenous. As suggested by Thomas (1971), species should be classified as autogenous or anautogenous based on local collections, and identified by dissecting individuals to see if nulliparous individuals have blood meals present in their gut. Ideally, individuals need to be dissected to be classified based on the gonotrophic stage at each sampling locality.

However, dissections were not completed.

Not all authors think that anautogeny is an adaptation for increased reproductive 49

output; rather they argue that it is tied to either survival or food reserves (Lang 1963). For example, Culex pipiens (Diptera: Culicidae) has autogenous and anautogenous forms, and the autogenous form can maintain greater lipid content from larval to adult forms than the anautogenous form(Rozeboom and Twohy 1958). The fourth prediction was supported, which lends support for the hypothesis that anautogenous adults would tend to be more

persistent and aggressive host seekers and hence, more likely caught around potential hosts.

There could be many reasons for why anautogeny exists as a successful life history strategy. Further research into larval habitat conditions, including rates of predation, as well as which species display which life history strategy, needs to be completed. Some species such as H. zonalis, which was thought to be anautogenous

(Thomas 1971), have subsequently been identified as facultatively autogenous

(McElligott and Lewis 1998), which argues for a range of strategies within a species, which might be locally adapted.

2 ln(1+n) Malaise catch

0 0 1 2 3 4 5 6

ln(1+n) sweep catch

FIGURE 4.1. Sweep and Malaise catches of anautogenous (open circles and dashed trendline) and autogenous (solid circles and solid trendline) Tabanidae species from

2011 in the western boreal forest of Ontario.

2 ln(1+n) Malaisecatch

0 0 1 2 3 4 5 6 ln(1+n) sweep catch

FIGURE 4.2. Sweep and Malaise catches of anautogenous (open circles and dashed trendline) and autogenous (solid circles and solid trendline) Tabanidae species from

2012 in the eastern boreal forest of Ontario.

TABLE 4.1. Classification of autogenous and anautogenous Tabanidae species collected in 2011 and 2012 in the Far North of

Ontario. Species are grouped based on listed authors' designations.* designates facultative autogeny.

Species Source Thomas McElligott and Leprince and Thomas (1971) Lewis (1998) Maire (1990) (1972)

Hybomitra hearlei X X X Hybomitra osburni X X Hybomitra pechumani X X Hybomitra frontalis X X X X Chrysops mitis X Chrysops frigidus X Hybomitra itasca X Chrysops nigripes X Chrysops zinzalus X Hybomitra liorhina X

Autogenous (not requiring blood) Hybomitra lurida X X* X Hybomitra zonalis X X* X

Hybomitra epistates X Hybomitra affinis X X Hybomitra arpadi X X X X Hybomitra illota X Hybomitra lasiophthalma X X Hybomitra nuda X X Hybomitra tetrica X Hybomitra typhus X X

Anautogenous (requires blood)

Chrysops excitans X 54

CHAPTER 5: GENERAL DISCUSSION

In this thesis, I set out three objectives. My first objective was to examine the species abundance, distributions and diversity in two poorly studied geographic areas in the Province of Ontario. This work was important because it will provide baseline data for further investigations, particularly in the climate of increased northern development. I then tested a hypothesis about the persistence of anautogeny in Tabanidae, based on their spatial and temporal distribution, and trap biases.

For the first and most important objective, I have increased the baseline distributional knowledge of Culicidae and Tabanidae. In the second chapter, I presented a manuscript which extends the northern range of 11 mosquito species and the southern range of one species out of the 19 species collected. In conjunction to extending known range information, I have filled gaps in range record information for 6 additional species and contributed the first record of Aedes pullatus in Ontario. In the third chapter, I presented my collections of 2168 tabanids. From these collections, I extended the known distribution of 10 Tabanidae species northward and the range of 3 species westward. I also presented the first record in Ontario of Hybomitra osburni and filled in range gaps of

16 other species. This is important because it indicates that there is still much to be 55

learned from studying this region. Existing distribution records are based on a few earlier studies. While extensive, these are nonetheless limited by the remoteness and the scale of sampling in northern Ontario.

The anticipated land use and habitat changes due to increased mineral extraction and climatic effects, argues for an urgency in increasing knowledge of species distribution in the boreal forest. Mosquitoes and tabanids are mainly blood-feeding

insects (anautogenous species) that therefore rely heavily on other species as hosts. They also have aquatic larvae therefore, they are susceptible to change in both aquatic and terrestrial habitats. They also have relatively rapid generation times compared to large mammals, so changes to populations could become evident quickly. There have not been any studies using Tabanidae to monitor environmental change, but it is possible that the way they could be used would be to monitor long term changes of the relative proportions of anautogenous and autogenous species within a habitat. However, in order for this to occur, studies of baseline information, such as this one, would need to be completed for the area being examined for change. This makes mosquitoes and tabanids useful as indicators of the effects of environmental and developmental change once baseline patterns of abundance have been documented.

The second objective was to find any evidence of a differential in west to east post-glacial migration, due to prevailing winds. I did not find any difference between species richness in the eastern boreal study sites, in 2012, and the western boreal study sites, sampled in 2011. More species were predicted in the eastern part of the province, based on the lognormal analysis of Culicidae and Tabanidae, 24 and 28 species in the east

and 15 and 26 in the west, respectively. However, in both culicids and tabanids, the 56

curves of the rarefaction analysis showed the same trend in both regions with error bars overlapping (Fig. 2.1 and Fig. 3.1), meaning that any difference in the number of species caught could be explained by sample size. This is an important effect that is often ignored. While many insect species, including Diptera, are known to be transported long distances by winds and weather patterns (Isard and Gage 2001), from my results, I cannot

support a biogeographic hypothesis of increased Tabanidae or Culicidae species diversity or richness in the eastern part of the boreal due to prevailing winds.

I also examined trapping biases for collection of tabanids. Since our sampling was conducted with both a passive Malaise trap and the active sampling sweep netting technique, I was able to determine the biases associated with each. In total, I collected 30 species of tabanids but the species composition in the resulting collections was different based on the sampling method. Sweep netting collected approximately 55% more individuals but fewer species, only 2 that were not collected in a Malaise trap. In total, the Malaise sampling collected fewer specimens than sweeping (850 compared to 1318), possibly because of a lack of a surveyor as a host, but collected more species, including 8 that were only collected with a Malaise trap. These findings are important because it shows that when collecting insect samples the purpose of determining species richness and diversity, multiple methods should be used to reduce inherent trap biases (Mihok et al. 2006). It is worth noting that for a survey of an understudied region, if the goal is documenting as many species as possibe, as was my goal, trapping methods should be selected based on success for collecting multiple taxa and should not target specific

species. For example, I acknowledge the success of CDC light traps in targeting 57

mosquitoes (Newhouse et al. 1966) or the NZI trap for biting flies (Mihok 2002), but due to the goal of mass multi-family collecting, over and above Culicidae and Tabanidae, these methods were not included. It should also be noted that collection sizes should be limited by the available resources, as proper specimen preparation and identification, although extremely important, can be time consuming when collections are large. For

example, a single summers’ collections (2 months) can take 6-8 months to pin, label and identify.

Many researchers rely solely on their preferred method of sampling, whether it be

NZI traps, Manitoba traps or malaise traps, because of the large amount of time and effort included in identifying large numbers of specimens. Including multiple methods would inherently increase the number of specimens and therefore the amount of time required to process the specimens. The significance of my results is that any survey of Tabanidae, and in general any insect surveys, must use a variety of sampling methods to overcome inherent trap biases. The general trend is that while one might be tempted to conclude that it is more efficient to use sweep netting to collect some species (e.g. anautogenous tabanids), to rely on a single collecting method in species diversity surveys is a mistake because it will produce a biased species composition.

The third objective was to test whether my trapping results were consistent with the idea of anautogeny in Tabanidae. The idea of trapping biases allowed me to analyze the persistence of blood-feeding in tabanids and the distributional variations between feeding types. Trapping methods revealed temporal variations in collections of

autogenous and anautogenous tabanids. I found evidence consistent with hypotheses 1 58

and 2 that anautogenous species would be collected over a longer period of time and that there was evidence for earlier emergence of anautogenous species. I was unable to find support for the third hypothesis that anautogenous larvae live in poor habitat, although I did not actually match species catches with the habitat from which they arose. It is possible that the expression of autogeny likely acts on a scale where in high quality habitat they are obligately autogenous and in marginal or poor habitats they are

facultatively autogenous. However, I was also able to find support for the fourth hypothesis that anautogenous species are more persistent and aggressive host seekers because they are, as expected, more frequently caught using the netting method, which occurred around potential hosts.

In general, my thesis provided a number of avenues for future research including tracking the changes in insect diversity within the Far North and studying the rate of larval development of anautogenous and autogenous tabanids.

LITERATURE CITED

Baxter, J.A., A.M. Mjeni and P.E. Morrison. 1973. Expression of autogeny in relation to

larval population density of Sarcophaga bullata Parker (Diptera: Sarcophagidae).

Canadian Journal of Zoology 51: 1189-1193.

Béla, T. 1995.Comparison of different methods for diversity ordering. Journal of

Vegetation Science 6:283-290.

Beckel, W. E. 1954. The identification of adult female Aedes mosquitoes (Diptera,

Culicidae) of the black-legged group taken in the field at Churchill, Manitoba.

Canadian Journal of Zoology 32: 324–330.

Beresford, D. V., and J. Sutcliffe. 2008. Sampling designs of insect time series data: are

they all irregularly spaced? Oikos 118: 115-121.

Beresford, D. 2011. Insect collections from Polar Bear Provincial Park, Ontario, with

new records. Journal of the Entomological Society of Ontario 142: 19–27.

Brewer, A. and Williamson, M. 1994. A new relationship for rarefaction. Biodiversity

and Conservation 3: 373-379.

Burger, J.F. 1977. The biosystematics of immature Arizona Tabanidae (Diptera).

Transactions of the American Entomological Society 103: 145-258. 60

Burnett, A.M. and K.L. Hays. 1974. Some influences of meteorological factors on flight

activity of female horse flies (Diptera: Tabanidae). Environmental Entomology 3:

515-521.

Butt, C., B. Hicks and C. Campbell. 2008. The diversity and abundance of Tabanidae in

black spruce forests and sphagnum bogs in Gros Morne National Park,

Newfoundland, Canada. Journal of the Acadian Entomological Society 4:7-13.

CESCC 2011. Wild species 2010: the general status of species in Canada. National

General Status Working Group. Canadian Endangered Species Conservation

Council. Available online at:

http://www.wildspecies.ca/wildspecies2010/downloads/wild-species-2010.pdf

Chao, A. and L. Jost. 2012. Coverage-based rarefaction and extrapolation: standardizing

samples by completeness rather than size. Ecology 93: 2533-2547.

Clements, A. N. 1992. The Biology of Mosquitoes: Volume 1. Development, nutrition,

and reproduction. Chapman and Hall, London UK. 509 pp.

Coulson, S.J., I.D. Hodkinson and N.R. Webb. 2003. Aerial dispersal of invertebrates

over high-Arctic glacier foreland: Midtre Lovénbreen, Svalbard. Polar Biology

26: 530-537.

Crins, W. J., Gray, P. A.,Uhlig, W. C. and Wester, M. C. 2009. The ecosystems of

Ontario, Part 1. Ecozones and ecoregions. Report SIBTER IMA TR-01. Ontario

Ministry of Natural Resources Inventory, Monitoring and Assessment Section.

Peterborough, ON. 71 pp.

Danks, H.V. 1992. Arctic insects as indicators of environmental change. Arctic 45:159-

166. 61

Darsie, R. F. and Ward, R. A. 2005. Identification and geographical distribution of the

mosquitoes of North America, North of Mexico. University of Florida Press,

Gainesville, FL. 400 pp.

Downes, J.A. 1958. The feeding habits of biting flies and their significance in

classification. Annual Review of Entomology 3: 249-266.

Downes, J. A. 1965. Adaptations of insects in the arctic. Annual Review of Entomology

10: 257–274.

Far North Science Advisory Panel. 2010. Science for a Changing Far North. The Report

of the Far North Science Advisory Panel. A report submitted to the Ontario

Ministry of Natural Resources. Queen’s Printer for Ontario, Toronto, ON. 141 pp.

Fukami, T., S. Naeem and D.A. Wardle. 2001. On similarity among local communities in

biodiversity experiments. Oikos 95: 340-348.

Gillis, M. D., A. Y. Omule, and T. Brierley. 2005. Monitoring Canada’s forests: the

national forest inventory. Forestry Chronicle 81: 214-221

Goldblatt, P., J.C. Manning, and P. Bernhardt. 2001. Radiation of pollination systems in

gladiolus (Iridaceae: Crocoideae) in southern Africa. Annals of the Missouri

Botanical Garden 88: 713-734.

Hongoh, V., L. Berrang-Ford, M.E. Scott and L.R. Lindsay. 2012. Expanding

geographical distribution of the mosquito, Culex pipiens, under climate change.

Applied geography 33: 53-62.

Hopper, K.R., P.H. Crowley and D. Kielman. 1996. Density dependence, hatching

synchrony, and within-cohort cannibalism in young dragonfly larvae. Ecology 77: 62

191-200.

Isard, S.A. and S.H. Gage. 2001. Flow of life in the atmosphere: An airscape approach to

understanding invasive organisms. Michigan State University Press. 240 pp.

Jenkins, D.W. and K.L. Knight. 1952. Ecological survey of the mosquitoes of southern

James Bay. American Midland Naturalist 47:456-468.

Krebs, C.J. 1989. Ecological Methodology. Harper and Row. 654 pp.

Lang, C.A. 1963. The effect of the temperature on the growth and chemical composition

of the mosquito. Journal of Insect Physiology 9: 279-286.

LePrince, D.J and A. Marie. 1990. Parity, stage of follicular development and sperm

presence in hemiarctic host-seeking Hybomitra populations (Diptera: Tabanidae).

Journal of Medical Entomology 27: 835-838.

Ludwig J.A and J.F. Reynolds. 1988. Statistical Ecology: a primer on methods and

computing. John Wiley & Sons, Inc. 368 pp.

Maire, A. 1984. Les Tabanidae (Diptera: Tabanidae) à la limite septentrionale des forêts

(domaine maritime du Nouveau-Québec). The Canadian Entomologist 116: 227-

233.

Maire, A. and J.P. Beaudoin. 1984. Hybomitra aequetincta et H. zonalis (Diptera:

Tabanidae): clé d’identification, distribution et abundance saisonnière au Québec-

Labrador. Revue d'Entomologie du Québec 29: 62-73.

McElligott, P.E.K. and D.J. Lewis. 1996. Development of larvae of Chrysops nigripes

Zetterstedt and Chrysops probably zinzalus Philip (Diptera: Tabanidae) in a

subarctic Labrador peatland. Canadian Journal of Zoology. 74: 1370-1375.

McElligott E.K. and Lewis D.J. 1996. Distribution and abundance of immature 63

Tabanidae (Diptera) in a subarctic Labrador peatland. Canadian Journal of

Zoology 74: 1364-1369.

McElligott, P.E.K. and D.J. Lewis. 1998. Seasonal changes in abundance and

gonotrophic age of host-seeking Tabanidae (Diptera) from subarctic Labrador

peatland. Journal of Medical Entomology. 35: 763-770.

Mihok, S. 2002. The development of a multipurpose trap (the Nzi) for tsetse and other

biting flies. Bulletin of Entomological Research 92: 385–403.

Mihok, S., D.A. Carlson, E.S. Krafsur and L.D. Foil. 2006. Performance of the Nzi and

other traps for biting flies in North America. Bulletin of Entomological Research.

96:387-397

Miller, L.A. 1951. Observations on the bionomics of some northern species of Tabanidae

(Diptera). Canadian Journal of Zoology 29: 240-263.

Muirhead-Thomson, R. C. 1991. Trap responses of flying insects: The influence of trap

design on capture efficiency. Academic Press, San Diego, CA. 304 pp.

Newhouse, V.F., R.W. Chamberlain, J.G. Johnston and W.D. Sudia. 1966. Use of dry ice

to increase catches of the CDC miniature light trap. Mosquito News. 26:30-35

Niemela, J., D. Langor and J.R. Spence. 1993. Effects of clear-cut harvesting on boreal

ground-beetle assemblages (Coleoptera: Carabidae) in western Canada.

Conservation Biology 7: 551-561.

OMNR (Ontario Ministry of Natural Resources). 2012. Wildlife research: Far North

biodiversity. Available online at:

http://www.mnr.gov.on.ca/en/Business/Wildlife/2ColumnSubPage/STDPROD_0 64

99955.html.

Pechuman, L. L., H. J. Teskey, and D. M. Davies. 1961. The Tabanidae (Diptera) of

Ontario. Proceedings of the Entomological Society of Ontario 91:77–121.

Raponi, M. 2014. Ectoparasitic influences of diptera on the habitat selection and behavior

of woodland caribou (rangifer tarandus) in a managed boreal forest landscape in

northern Ontario. M.Sc Thesis. Trent University, Peterborough, Ontario

Ringrose, J.L, K.F. Abraham, D.V. Beresford. 2013. New range records of mosquitoes

(Diptera: Culicidae) from northern Ontario. Journal of the Entomological Society

of Ontario 144: 3-14.

Rozeboom , L.E. and D.W. Twohy. 1958. Nutritive reserves in males of autogenous and

anautogenous populations of Culex pipiens. The Journal of Parasitology 44: 422-

424.

Ruxton, G. D. 2006. The unequal variance t-test is an underused alternative to Student's t-

test and the Mann–Whitney U test. Behavioral Ecology 17: 688-690.

Saska, P., W. van der Werf, L. Hemerik, M.L. Luff, T. Hatten and A. Honek. 2013.

Temperature effects on pitfall catches of epigeal arthropods: a model and method

for bias correction. Journal of Applied Ecology 50: 181-189.

Snyder, T.E. 1916. Notes on horse flies as a pest in southern Florida. Proceedings of the

Entomological Society of Washington 19: 208-211.

Spielman, A. 1971. Bionomics of autogenous mosquitoes. Annual Review of Entomology

16: 231-248.

Steward, C. C. and McWade, J. W. 1960. The mosquitoes of Ontario (Diptera: Culicidae)

with keys to the species and notes on distribution. Proceedings of the 65

Entomological Society of Ontario 91: 121–188.

Taylor, L.R. 1963. Analysis of the effect of temperature on insects in flight. Journal of

Animal Ecology 32: 99-117.

Teskey, H. J. 1990. The horse flies and deer flies of Canada and Alaska (Diptera:

Tabanidae). The Insects and Arachnids of Canada, Part 16. Agriculture Canada.

Publication 1838. 381 pp.

Thielman, A.C, and Hunter, F.F. 2007. Photographic Key to the Adult Female

Mosquitoes (Diptera: Culicidae) of Canada. Canadian Journal of Arthropod

Identification No. 4, 14 December 2007. Available online at

http://www.biology.ualberta.ca/bsc/ejournal/th_04/th_04.html, doi:

10.3752/cjai.2007.04.

Thomas, A.W. 1971. Autogeny and anautogeny in some species of tabanids (Diptera:

Tabanidae) in Alberta, Canada. PhD thesis, University of Alberta, Edmonton.

Thomas, A.W. 1972. Physiological age structure of adult tabanids populations (Diptera:

Tabanidae) in Alberta, Canada. Journal of Medical Entomology 9: 295:300.

Thomas, A.W. and Marshall. S.A. 2009. Tabanidae of Canada, east of the Rocky

Mountains 1: a photographic key to the species of Chrysopsinae and Pangoniinae

(Diptera:Tabanidae). Canadian Journal of Arthropod Identification No. 8, 25 June

2009. Available online at http://www.biology.ualberta.ca/bsc/ejournal/tm_08html,

doi:10.3752/cjai.2009.08.

Thomas, A.W. 2011. Tabanidae of Canada, east of the Rocky Mountains 2: a

photographic key to the genera and species of Tabanidae (Diptera: Tabanidae).

Canadian Journal of Arthropod Identification No.13, 16 February 2011, Available 66

online at http://www.biology.ualberta.ca/bsc/ejournal/t_13/t_13.html,

doi:10.3752/cjai.2011.13.

Thompson, P. 1969. Collecting methods for Tabanidae (Diptera). Annals of the

Entomological Society of America 62: 50-57.

Turell, M.J., O’Guinn, M.L., Dohm, D.J. and Jones, J.W. 2001. Vector competence of

north America mosquitoes (Diptera:Culicidae) for West Nile Virus. Journal of

Medical Entomology 38: 130-134.

Van Buskirk, J. 1989. Density-dependent cannibalism in larval dragonflies. Ecology. 70:

1442-1449.

Vockeroth, J. R. 1954. Notes on the identities and distributions of Aedes species of

northern Canada, with a key to the females (Diptera: Culicidae). The Canadian

Entomologist 6: 241–255.

Warner, B.G. and Asada, T. 2006. Biological diversity of peatlands in Canada. Aquatic

Sciences 68: 240-253.

Wild Species. 2010. Wild Species 2010 Report, in Wild Species, the general Status of

Species in Canada. Available online at:

http://www.wildspecies.ca/home.cfm?lang=e.

Wood, D.M., Dang, P.T. and Ellis, R.A. 1979. The mosquitoes of Canada (Diptera:

Culicidae). The Insects and Arachnids of Canada, Part 6: Agriculture Canada.

Publication 1686. 390 pp.

Wood, D.M. 1985.Biting flies attacking man and livestock in Canada. Agriculture 67

Canada. Publication 1781/E. 40 pp.

WRBU (Walter Reed Biosystematics Unit). 2013. Traditional Mosquito Classification,

July 2013. Available online at:

http://www.wrbu.org/docs/mq_ClassificationTraditional201307.pdf.

APPENDICES

APPENDIX 1: SITE DESCRIPTIONS

APPENDIX TABLE 1. Descriptions of 2011 and 2012 sample sites. Definitions are based on OPIAM (Ontario Parks Inventory and

Monitoring Program: Guidelines and Methodologies-Version 1.4 Draft, 25 May 2012)

Sampling Community Plant Cover Year Sampling Dates Latitude Longitude Ecological System Type Dominant Plant Form Class May 31-June 7 54° 25' 50" 89° 40' 43" Terrestrial Forest Shrub Low Shrub May 31-June 7 53° 12' 8" 89° 6' 28" Terrestrial Forest Tree-Coniferous Open Treed June 8-15 54° 9' 30" 92° 1' 44" Terrestrial Forest Tree-Deciduous Low Treed June 8-15 53° 45' 35" 88° 54' 51" Wetland Bog Tree-Coniferous Sparse Treed June 16-June 23 54° 28' 19" 88° 33' 33" Terrestrial Forest Bryophytes Herbaceous June 16-June 23 54° 27' 1" 90° 21' 38" Terrestrial Forest Tree-Coniferous Open Treed 2011 June 28-July 5 53° 27' 40" 93° 2' 33" Terrestrial Forest Tree-Coniferous Open Treed Sparse Low June 28-July 5 52° 49' 28" 94° 13' 38" Terrestrial Meadow Tree-Coniferous Treed Aquatic/Terrestrial July 6-July 13 52° 27' 37" 91° 49' 9" Interface Forest/Lake Shrub Low Shrub July 6-July 13 53° 36' 9" 93° 32' 10" Terrestrial Forest Tree-Coniferous Closed Treed July 14-July 21 53° 44' 12" 92° 46' 4" Wetland Bog Tree-Coniferous Sparse Treed Aquatic/Terrestrial June 5-June 11 52° 46' 35" 81° 57' 48" Interface Marsh Graminoid Herbaceous Sparse Low June 5-June 11 51° 55' 53" 82° 39' 13" River Terrestrial Shrub Shrub 2012 Sparse Low June 12-June 18 51° 26' 40" 80° 23' 11" Wetland Meadow Graminoid Shrub June 12-June 18 51° 39' 8" 81° 50' 57" Terrestrial Forest Tree-Mixed Closed Treed June 19-June 25 51° 58' 8" 81° 39' 23" Aquatic/Terrestrial River Tree-Coniferous Open Treed

Interface June 19-June 25 52° 53' 25" 82° 41' 2" Terrestrial Meadow Graminoid Non-vascular Aquatic/Terrestrial Mixed June 26-July 2 52° 28' 27" 82° 49' 2" River Interface Herbs/Submergent Herbaceous Aquatic/Terrestrial June 26-July 2 51° 29' 53" 83° 17' 24" Interface Forest/River Mixed Herbs Herbaceous Aquatic/Terrestrial July 3-July 9 51° 47' 50" 83° 2' 25" River Interface Shrub Low Shrub Sparse Low July 3-July 9 52° 23' 20" 82° 8' 2" Terrestrial Forest Shrub Treed Aquatic/Terrestrial Sparse July 10-June 16 52° 18' 23" 83° 22' 44" River Interface Mixed Herbs Herbaceous July 10-July 16 51° 21' 22" 82° 3' 24" Wetland Fen Shrub Low Shrub

APPENDIX 2: SITE PICTURES

APPENDIX FIGURE 1. Typical Malaise trap setup in a fen bordered by a black spruce

(Picea mariana) forest.

APPENDIX FIGURE 2. Typical camp setup showing malaise trap placement in a fen bordered by a black spruce (Picea mariana) forest.

APPENDIX FIGURE 3. Typical camp setup in a bog. Malaise trap was placed behind the photographer.

APPENDIX FIGURE 4. Typical camp setup in a bog bordered by a black spruce (Picea mariana) forest.