National Library Bibliothèque nationale of Canada du Canada
Acquisitions and Direction des acquisitions ct Bibliographie Services Branch des services bibliographiques
395 Wetlt0Ç11on StrCCl 395. ru~ Wcll1nqlon Otlawa. Ontario Ottawa (OnlalMJ) K1AQN4 K1AON4 '.\.1 '.~. \ , ....,. ",~""",,'.,
,',' loi" ...... ~' •• ".r,.".....,. NOTICE AVIS
The quality of this microform is La qualité de cette microforme heavily dependent upon the dépend grandement de la qualité quality of the original thesis de la thèse soumise au submitted for microfilming. microfilmage. Nous avons tout Every effort has been made to fait pour assurer une qualité ensure the highest quality of supérieure de reproduction. reproduction possible.
If pages are missing, contact the S'il manque des pages, veuillez university which granted the communiquer avec l'université degree. qui a conféré le grade.
Sorne pages may have indistinct La qu~;ité d'impression de print especially if the original certaines pages peut laisser à pages were typed with a poor désirer, surtout si les pages typewriter ribbon or if the originales ont été university sent us. an inferior dactylographiées à l'aide d'un photocopy. ruban usé ou si l'université nous a fait parvenir une photocopie de qualité inférieure.
Reproduction in full or in part of La reproduction, même partielle, this microform is governed by de cette microforme est soumise the Canadian Copyright Act, à la Loi canadienne sur le droit R.S.C. 1970, c. C-30, and d'auteur, SRC 1970, c. C-30, et subsequent amendments. ses amendements subséquents.
Canada • ASPEcrs OF THE BIOLOGY OF HORSE FLIFS AND DEER FLIFS (Diptera: Tabanidae) IN SUBARcnC LABR>_'nR: LARVAL DISTRIBUTION AND DEVELOPMENT, BIOLOGY OF HOST-SEEKING FEMALES, AND EFFECT OF CLIMATIC FACTORS ON DAlLY ACTIVITY.
By
Paul Edward Kaye McElligott, B.Sc., M.Sc.
Department ofEntomology, McGill UniveISÏty • Ste-Anne-de-Bellevue, Quebec
AThesis Presented to the FacilJ.ty ofGraduate Studies and Research -'" in Partial Fnlfillment ofthe Requirements ofthe Degree of Doctor ofPhilosophy
C P. McElligon August, 1992 • Nationallibrary Bibliothèque nationale ..... of Canada du Canada Acquisitions and Direction des acquisitions el Bibliographie Services Branch des services bibliographiques
395 Welh~ton Street 3SS. rue Wt:illnglon Ottawa.Ontano Onawa (OnI~HtO) K1AON4 K1AON4
The author has granted an L'auteur a accordé une licence irrevocable non-exclusive licence irrévocable et non exclusive allowing the National Library of permettant à la Bibliothèque Canada to reproduce, loan, nationale du Canada de distribute or sell copies of reproduire, prêter, distribuer ou hisjher thesis by any means and vendre des copies de sa thèse in any form or format, making de quelque manière et sous this thesis available to interested quelque forme que ce soit pour persons. mettre des exemplaires de cette thèse à la disposition des personnes intéressées.
The author retains ownership of L'auteur conserve la propriété du the copyright in hisjher thesis. droit d'auteur qui protège sa Neither the thesis nor substantial thèse. Ni la thèse ni des extraits extracts from it may be printed or substantiels cie celle-ci ne otherwise reproduced without doivent être imprimés ou hisjher permission. autrement reproduits sans son autorisation.
ISBN 0-315-87947-5
Canada •
Suggested Short Title:
Aspects of the Biology of Horse Flies and Decr Flies in • Subarctic Labrador.
• Il
Abstract
• Ph.D. Paul McElligott Entomology
Aspects of the Biology of Horse Flies and Deer Flies (Diptera: Tabanidae) in
Subarctic Labrador: Larval Distribution and Development, Biology of Host
Seeking Females, and Effect of Climatic Factors on Activity.
Larval tabanids were colleeted twice weekly from eight locations in Iron
Ann fen, a peatland in subarctic Labrador near Schefferville, Quebec, June
through August, 1990 and 1991. Ofthe 476 tabanid larvae colleeted, 827% were
Chrysops (5 spp.), 17.0% were Hybomitra (5 spp.), and 03% were Atylotus • sphagnicola Teskey; the mest abondant species in the fen were C. zinzalus Philip (31%), and C. nigripes ZetteIStedt (24%). Species- and genera-specific
microhabitat preferences were apparent; in general Chrysops spp. preferred drier
regions of the fen than did Hybomitra spp.. Larvae of C. zinzalus and C. nigripes
appear to require 3-4 yeaxs to complete thëir larval development in subarctic
regions, based upon their patterns ofseasonal growth. Annual variation in larval
year-class sizes suggests that populations vaI} dramatically within individual
peatlands from year 10 year.
Adult hOISe flies and deer flies were colleeted using canopy and Malaise
traps at two locations in the Schefferville area, Iron Ann fen and Capricom fen,
from late June until early August in 1990 and 1991. Seventeen tabanid species • were colleeted, six Chrysops spp., 10 Hybomitra spp., one Atylotus sp.; Hybomitra ili spp. comprised 96% of collections. Adult abundance of different species varied • markedly between the two study sites; in general Iron Arm fen had a more abundant and diverse tabanid fauna than Capricom. Of the Hybomitro spp.
common in the Schefferville area, H. lurida (Fallen), H. aequetincta (Becker), H.
arpadi (Szilady), and H. zonalis (Kirby) were most abundant in early to mid July,
whereas H. heariei (Philip) and H.pecJuunani Teskey and Thomas were common
from mid July until early August. Chrysops spp. were most active in late July and
early August; Chrysops nigripes and C. zinzaIus were the most common specie::.
For each of 10 tabanid species, samples of 10 fiies were taken from daily
trap catches for dissection and determination ofparity, yolk deposition, and fat
body deposition. In the Schefferville area, H. arpadi and H. aequetincta are
obligately anautogenous (i.e., require a blood-meal in order to mature eggs), H.
• lurida and H. zonalis are facultatively autogenous (i.e., capable of maturing eggs
without in some cases), and H. pechumani. H. heariei. H.frontalis (Walker). H.
astuta (Osten Sacken), C. zinzaIus and C. nigripes are obligately autogenous (i.e.
always mature a fust batch of eggs without taking blood). Based upon
gonotrophic age-grading of nulliparous individuaIs, the majority of H. aequetineta
and H. arpadi females emerge either at the beginning of the flight season, midway
through the season, or both, depending upon year and site. Most H. zonalis
emergê midway through the fiight season. Nulliparous female tabanids of
anautogenous or facultatively autogenous species usually carry considerable • amounts offat body within their abdomens. This fat is depleted as nutrients are iv tIansferred to the developing oocytes; recenùy parous females cany very lime fat • body. The eITeet of meterological variables on tabanid daily activity was
investigated using a canopy tIap incorporating an eleetronic insect counter, a
computerized data-Iogger, and sensors to measure air temperature, solar radiation,
wind speed and direction, and relative humidity. Tabanid numbers and
meteorological variables were recorded every halfhour between 0530 and 2130
EDT. Tabanid activity only occurred at temperatures exceeding 90 C and levels of
2 solar radiation exceeding approximately 10 W1m • The onset of tabanid activit'J
in the morning was usually temperature dependant, whereas the cessation of
activity in the evening was light dependent. Tabanid activity increased with
increasing temperature between 10 and 190 C. Multiple regression analyses
• determined that during the daylight hours temperature, relative humidity, and
date offlight season most significanùy affect level oftabanid host-seeking activity.
• v Résumé • Ph.D. Paul McElligott Entomology
Aspects de la Biologie du Taon (Diptère: Tabanidae) au Labrador Sous-Arctique:
Distribution et Développement Larvaire, Biologie et Activité de Recherche-Hôte
de la Femelle, et Effets des Facteurs Climatiques sur l'Activité Journalière.
Des larves de taon furent collectées deux jours par semaine, de juin à août
1990 et 1991, de huit sites d'Iron Arm, une tourbière du Labrador sous-arctique,
près de Schefferville au Québec. Des 476 larves collectées, 82.7% appartenaient
au genre Chrysops (5 espèces), 17.0% au genre Hybomitra (5 espèces) et 0.3%
étaient des Atylotus sphagnicola Teskey. Les espèces les plus abondantes dans la • tourbière étaient C zinzalus Philip (31%) et C nigripes Zetterstedt (24%). Les préfèrences en microhabitates étaient apparentes entre les différents genres et
espèces; comparées aux Hybomitra spp., C/zrysops spp. préfèraient, généralement,
les régions les plus séches de la tourbière. Les larves de C zinzalus et C nigripes
semblent mettrent 3 à 4 ans pour compléter leur développement larvaire dans les
regions sous-arctiques et ceci sur base de leur mode de croissance saisonnière.
Les variations annuelles de la taille des cohortes larvaires, suggérent que les
populations varient fortement d'année en année dans les tourbières même.
Des taons adultes furent attrapés à l'aide de pièges "Canopy" et de tentes
"Malaise", à deux endroits dans les regions de Schefferville, au tourbières d'Iron • Arm et de Capricorn, entre fin juin et début août 1990 et 1991. Dix-sept espèces vi de taon furent collectées, dont: six Chrysops, 10 Hybomitra, et un Atylotus; • Hybomitra spp. constituaient 96% de la collecte. L'abondance des adultes des différentes espèces varaient entre les deux sites étudiés; en general, la tourbière
d'Iron Arro avait une faune de taon plus abondante et plus diversifiée que
Capricorn. De toutes les espèces d'Hybomitra fréquents dans la region de
Schefferville, H. lurida (Fallen), H. aequetincta (Becker), H. arpadi (Szilady), et H.
zonalis (Kirby) abondaient au début à la mi-juillet, alors que H. hearlei (Philip) et
H. pechumani Teskey and Thomas étaient fréquentes de la mi-juillet.au début
d'août. Chrysops spp. étaient actifs surtout du fin juillet au début d'août; C/zrysops
nigripes et C zinzalus étaient les plus abondantes.
De chacune des 10 espèces de taons atttapées dans les pièges, des
échantillons de 10 mouches furent préléres et disséqués afin de déterminer la • parité, le dépôt du vitellus et du corps gras. Dans la region de Schefferville, H. arpadi et H. aequetincta sont "obligatoirement anautogenous" (i.e., on besoin d'une
ingestion de sang afin d'amener ,leurs oeufs à maturité), H. lurida et H. zonalis
sont des "autogenous facultatifs" (i.e., capable de former des oeufs mûres sans
ingestion de sang dans certains cas), et H. pechumani, H. hearlei, H. frontalis
(Walker), H. astuta (Osten Sacken) C zinzalus et C nigripes sont des "autogenous
obligatoires" (i.e., portent toujours à maturité un premier lot d'oeufs sans ingestion
de sang). En se basant sur l'age reproductif des individus nullipares, on remarque
que la majorité des femelles de H. aequetincta et H. arpadi émergent ai.l3ébut
et/ou au milieu de la saison de vol selon l'année et le site. La plupart de H. • zonalis émerge à mi-saison de vol. Les taons femelles nullipares des espèces V11 "anautogenous" ou "autogenous facultatives· possédent une importante quantité de • corps gras dans leurs abdomens. Ce gras est épuisé par le transfert de éleménts nutritifs aux oocytes en développement; les femelle de parité récente possédent
trés peu de corps gras.
L'ffet des paramètres météorologiques sur l'activité journaliére des taons
fût étudié à l'aide d'une piège "Canopy" comportant un compteur d'insectes
éléetronique, un systéme d'acquisition de données géré par ordinateur, et des
sondes pour mesurer la température de lair, la radiation solaire, la vitesse et la
direction du vente et l'hutnidité relative. La nombre de taons et les paramètres
météorologiques furent notées à chaque detni-heure entre 0530 et 2130. Les
taons étaient actifs uniquement lorsque la température dépassait 9· C et la
2 radiation solaire 0 W1m • Le matin, le diclenchement de l'activité des taons • dépendait de la température alors que le soir, l'arrêt d'activité dépendait de la lumiére. Entre 10 et 19· C, l'activité des taons augmentaient avec la température.
Des analyses de regression multiples on démontré que durant le jour, la
température, l'hutnidité relative et la date de la saison de vol ont l'effet le plus
significantif sur le niveau d'activité de recherche l'hôte du taon.
• viii TABLE OF CONTENTS
• LIST OF TABLES ...... ••.....•...... •...... •.•...... xi i LIST OF APPENDICES xi i i LIST OF FIGURES •...... ••.. xiv ACKNO"lLEGEMENTS ..•..••...... •...... •...... •.•..•..•••••.••.•.••• xvi i FOREWORD .....•.•••.•...••..••••.....••••••....•...••••...•••.•••••• xi x CHAPTERS: 1. Distribution and Development of Larval Tabanidae (Diptera) in a Subarctic Peatland
I. LITERATURE REVIEW ..••...•••••...... •.•.•••....•.••..••.•.••••. 2 1. Biology of Immature Tabanidae ••••••....••••...•••••...••.. 2 1.1. Li fe Cycle ••••..•••••••..•.••••••.•.•••.••.•••.•• 3 1.1.1. Egg Stage •.••••••••••.•••••.•••.••.••••••. 3 1.1.2. Larval Stages ••..•.•••.•..•.•.•.•..•.•••.. 6 1.1.3. Pupal Stage •.••••••••..••••....••••••••••• 10 1.2. Larval Diet •••..••.••.••.•.•...••.....••••.•.•.•. Il 1.3. Larval Habitats ...... •..•.•.•••....•.•••••.••••. 12 1.4. Larval Densities ••...•..•••.•.••••.•...••.•••.••• 14 1.5. Natural Enemies of Immature Tabanidae .•...... ••.. 15 • 2. Immature Tabanidae in Subarctic Canada .....•...... •.... 17 1I. INTRODUCTION ...... •....•.•••....•....•..•.....•. 24 III. MATERIALS AND METHODS ..••....•••...•..•.•..••••..•..••.•..•••• 25 1. Study Area ••••.....••...••.••....•.•.•••..•.••.••.•...••.• 25 2. Sampl i ng ..•••••••••.•..••...•.•.....••••••.•..•••0...... 26 2.1. Wet Extraction ••. : •.••••...•.•..•.••.••••.•.••••. 26 2.2. Dry Extraction .•••...•••.....••...•.•..•.•..•.•.. 27 2.3. Emergence Traps ••...... ••.•.....••.•...•.•.•.•..• 2B 3. Preparation and Treatment of Larvae •.•.•...••.•••.•.•••••. 29 IV. RESULTS AND DISCUSSION ...••...•..•.••.•.•..••••.•••.•....•••.• 30 1. Larval Diversity •...••...... ••..•.. : .•..••.•••..•.•.•.•• 30 1.1. Chrysops spp. . •••.•..•.. ~ .•.•...•••••••••.•..•••. 30 1.2. Hybomitra spp .•.•....•••....••.•.•.•.•..•.•.•.•.. 32 2. Pupae ••••••.•.•••••.•.••..••••.•.••.•..••••••••.•.•.•.••.• 32 3. Eggs •.•••••.•.••••••••••..•••..•••.••••••••.••.••.•.•••.•• 35 4. Habitat Preferences 35 4.1. Chrysops spp. ..•••...••••.••••..•.•..••••.••••••• 35 4.2. Hybomitra, Atylotus spp •.•.•.•••••••••.•••••.••.• 36 • 4.3. Emergence Traps •.•.••••••••••••..•.•••••••••••••• 37 IX 5. Seasonal Patterns of Larval Growth 38 5.1. Chrysops nigripes and C. zinza7us 39 5.2. Other Chrysops spp. .. 42 • 5.3. Summary: Chrysops Life Cycles 43 5.4. Annual variation in Chrysops Larval Cohorts 44 5.5. Hybomitra spp 46 5.6. Growth of Captive Chrysops and Hybomitra Larvae .. 47 V. LITERATURE ClTED 48
2. Diversity, Seasonal Activity, and Seasonal Changes in the Gonotrophic Age Structure of Host-Seeking Tabanidae From Two Peatlands Near Schefferville, Quëbec. I. LITERATURE REVIEW .•.••...... •...... 72 1. Distribution of Adult Tabanids 72 1.1. Larval Habitat vs. Adult Distribution 72 1.2. Habitat Selection by Host-Seeking Females 73 1.3. Mating Site Selection by Males 73 1.4. Dispersal 74 1.4.1. Interspecific Differences in Dispersal ..• 75 1.4.2. Effect of Topography on Dispersal 75 1.4.3. Effect of Host Distribution on Dispersal. 76 1.4.4. Effect of Weather of Dispersal 77 1. 4.5. Di spersal Di stance ...... •...... 78 1.5. Resting Behavior 79 2. Gonotrophic Development of Female Tabanids 80 • 2.1. Determination of Follicular Development 80 2.2. Determination of Parity 82 2.3. Autogeny and Anautogeny 83 2.3.1. Autogenous Species 84 2.3.2. Anautogenous Species ...•...... •...... 85 2.4. Timing of Ovarian Development in Tabanids 86 3. Tabanid Research in Northern Canada ...... •. 87 3.1. The Tabanid Fauna of Northern Canada 87 3.2. Seasonal Activty of Northern Tabanids 88 3.3. Other Research on Tabanids in Northern Canada 88 II. INTRODUCTION 91 III. MATERIALS AND METHODS 92 1. Study Areas 92 2. Collection of Adult Tabanids 93 3. Categorization of Ovarian Development 94 3.1. Dissection •...... : 95 3.2. Determination of Parity : 95 3.3. Gonotrophic Age-Grading 95 4. Categorization of Fat Body Deposition 96 • x IV. RESULTS AND DISCUSSION 97 1. Faunal Di vers ity 97 1.1. Faunal Diversity in the Schefferville Area 97 • 1.1.1. Tabaninae 97 1.1.2. Chrysopsinae 100 1.2. Local Faunal Diversity 101 1.2.1. Tabaninae 101 1.2.2. Chrysopsinae 102 1.2.3. Uncommon and Rare Species 103 1.3. Factors Affecting Faunal Diversity •...... 103 1.3.1. Geographie Variation in Faunal Diversity. 103 1.3.2. Annual Variation in Faunal Diversity 105 2. Tabanid Flight Seasons in the Schefferville Area 106 2.1. Host-Seeking Females 106 2.2. Males 108 3. Seasonal Variation in Gonotrophic Age of Host-Seeking Females 109 3.1. Obligately Autogenous Species 110 3.1.1. Hybomitra aequetincta ...... •.....•..•. 110 3.1.2. Hybomitra arpadi •.•.•.....•.•.••.....••• 115 3.2. Facultatively Autogenous Species ...•..••.•....•. 115 3.2.1 Hybomitra 7urida •••....•.•••...... •. 116 3.2.2. Hybc~itra zona7is .....•.....•...... •• 116 3.3. Obligately Autogenous Species ...... •...... 118 3.4. Other Speci es •...... 118 4.' Fat Body Deposition and Utilization 119 • V. LITERATURE CITED ...... •...... •...... 122 3. The Effect of Selected Climatic Factors on the Daily Activity of Tabanids (Diptera) Collected at a Peatland in Subarctic Labrador. 1. LITERATURE REV 1EW ...... •.••...... •...... •••.•. 149 1. Climatic Factors and Tabanid Host-Seeking Activity 149 1.1. Temperature 149 1.2. Light Intensity 150 1.3. Wind Speed 150 1.4. Barometric Pressure and Other Factors 151 1.5. Multiple Regression Analyses ...•...... 152 2. Interspecific Variation in Tabanid Host-Seeking Activity . 152 __ II. INTRODUCTION...... ••...... • 154 III. MATERIALS AND METHODS .•.•...... •.•••.•.•.•.•...... 154 1. Study Area •...... •...... •..••.•••••••••.•...... ••• 154 2. Data Collection Methods ...... •. 155 3. Data Analyses ...... •...... ::;::156 3.1. ' Threshold Values of Temperature and Solar Radiation ...... ••.•.•.••.•. 156 • 3.2. Multiple Regression Analyses ...... ••.•.•.. 156 XI IV. RESULTS AND DISCUSSION .•.•.••••..•...•.••.•••.•.•.•.••.....•. 157 1. Thresho1ds 158 2. Regressions 159 • 3. The Use of Data-Loggers in Future Studies 161 V. LITERATURE CITED •••••••••••••..•••.••.•••.•.••...... •...... 162
GENERAL CONCLUS IONS ••••••.••••••••••..•••••••..••.•.••••.•.••....•• 172
•
,- • Xli LIST OF TABLES • Chapter l. Table 1. Recorded lengths of tabanid life cycles in the laboratory and field 22 Table 2. Recorded lengths of tabanid pupation periods in the l aboratory 23 Table 3. Average percent cover of major vegetation and exposed substrate types, relative wetness, pH, and years sampled at eight sampling sites in Iron Arm fen, 1990 and 1991 54 Table 4. Sampling effort and total number of larvae collected at eight sampling sites in Iron Arm fen, 1990 and 1991 ..•.•...... 55
Chapter 2. Table 1. Species, relative abundances, and dates of observed flight seasons of tabanid species collected in 1990 and 1991 from Iron Arm and Capri corn fens, near Schefferville, Quebec .. 131
Chapter 3. Table 1. Time, temperature, and solar radiation at onset and cessation • of tabanid activity, Iron Arm fen, July 20 - August 5, 1991 ...... •...... •...... •..•.... 164 Table 2. Regression equations for meteorological factors affecting ho st seeking activity of adult horse flies at Iron Arm fen, Labrador '...... •...... •••..•••...... •.....•. •. 165
• XII\ • LIST OF APPENDICES Appendix 1. Numbers of Hybomitra and Chrysops spp. dissected during each week and each day of the flight seasons of 1990 and 1991 at (a) Capri corn fen, (b) Iron Arm fen 147
•
• xiv LIST OF FIGURES • Chapter l Fig. l. Location of Schefferville, Quebac, and approximate location of study area ...... •...... 56 Fig. 2. Map of Iron Arm fen, 20 km ne Schefferville, Quebec, indicating moisture regimes and locations of sampling sites in 1990 (Site 1-4) and 1991 (Sites 1-8) 57 Fig. 3. Mean (± standard error) weekly temperature in top 10 cm of substrate in Iron Arm fen during summers of 1990 and 1991 •.•••...•...•.•••...•..•...•....•.•...... •...... 58 Fig. 4. Schematic diagram of apparatus used in (a) wet extraction, and (b) dry extraction ...... •.•••..•.•.•.....•.•....•.•. 59 Fig. 5. Patterns of pubescence on larvae of (a) Chrysops presumed zinza7us Philip, (b) C. nigripes Osten Sac ken variant collected at Schefferville, and (c) C. nigripes described by Teskey (1969) 60
Fig. 6. Weekly mean (± standard error) lengths of larval Chrysops nigripes Osten Sacken in three size categories collectee from all sites in Iron Arm fen in 1990 and 1991 •....•....•....• 61 Fig. 7. Weekly mean (± standard error) lengths of larval Chrysops • presumed zinza7us Philip in three size categories collected from all sites in Iron Arm fen in 1990 and 1991 •••...... • 62 Fig. 8. Schematic representation of presumed life cycle of Chrysops nigripes Osten Sacken in Iron Arm fen, indicating likely growth patterns of 1986-1991 larval cohorts •...... 63 Fig. 9. Schematic representation of presumed life cycle of Chrysops zinza7us Philip in Iron Arm fen, indicating likely growth patterns of 1986-1991 larval cohorts ...... •.... 64
Fig. 10. Weekly mean (± standard error) lengths of larval Chrysops ater Macquart in three size categories collected from all sites in Iron Arm fen in 1990 and 1991 ....•.•, ..•....•...... 65
Fig. Il. Weekly mean (± standard error) lengths of larval Chrysops frigidus Osten Sacken in three size categories collected from all sites in Iron Arm fen in 1990 and 1991 ...... •.•• 66 Fig. 12. Weekly mean (± standard error) lengths of larval Chrysops furcatus Walker in three size categories collected from all sites in Iron Arm fen in 1990 and 1991 ••.•.•...... 67 Fig. 13. Lengths of larval Hybomitra arpadi Szilady collected from all • sites in Iron Arm fen in 1990 and 1991 ...... •...... • 68 xv Fig. 14. Percentage of Chrysops nigripes Osten Sac ken larvae from Iron Arm in 1990 and 1991 fen which fell into each of three size • categories 69 Fig. 15. Percentage of Chrysops presumed zinza7us Philip larvae from Iron Arm fen in 1990 and 1991 fen which fell into each of three size categories 70
Chapter 2. Fig. 1. Map of vicinity of Schefferville, Quebec, indicating locations of Iron Arm fen and Capri corn fen 132 Fig. 2. Seasonal abundance of Hybomitra spp. at Iron Arm fen in 1990 and 1991 ...... •.....••.•...... 133 Fig. 3. Seasonal abundance of Hybomitra spp. at Capri corn fen in 1990 and 1991 ...... •...... •...... 134 Fig. 4. Seasonal abundance of Chrysops spp. at Iron Arm fen in 1990 and 1991 .....•.•••....•..••••••...••...... •...•.•...... 135 Fig. 5. Seasonal abundance of Chrysops spp. at Capri corn fen in 1990 and 1991 ...... •...... 136 Fig. 6. Proportions of host-seeking female Hybomitra aequetincta collected at weeks 1-6 of the flight season with ovarioles at • successive stages of development 137 Fig. 7. Average numbers of host-seeking female Hybomitra aequetincta collected during weeks 1-6 of its flight season with Nulliparous Stage l, Nulliparous Stage II, and Parous ovarioles ••..•.....•..•.••....•••...... •••••...... 138 Fig. 8. Proportions of host-seeking female Hybomitra arpadi collected at weeks 1-6 of the flight season with ovarioles at successive stages of development 139 Fig. 9. Average numbers of host-seeking female Hybomitra arpadi collected during weeks 1-6 of its flight season with Nulliparous Stage l, Nulliparous Stage II, and Parous ovarioles ...•.... , ..••...... •....•...... 140 Fig. 10. Proportions of host-seeking female Hybomitra 7urida collected at weeks 1-6 of the flight season with ovarioles at successive stages of development .•..••...... •...... •.•...... 141 Fig. Il. Proportions of host-seeking female Hybomitra zona7is collected at weeks 1-6 of the flight season with ovarioles at successive • stages of development ..•....•...... •...... •...... 142 X"VI Fig. 12. Average numbers of host-seeking female Hybomitra arpadi collected during weeks 1-6 of its flight season with Nulliparous Stage 1, Nulliparous Stage II, and Parous • ovarioles collected per week . 143 Fig. 13. Proportions of host-seeking females of obligately anautogenous species (i.e. Hybomitra aequetincta and H. arpadi) at successive stages of ovarian ùevelopment with abdomens containing four different quantities of fat body 144 Fig. 14. Proportions of host-seeking females of facultatively autogenous species (i.e. Hybomitra 7urida and H. zona7is) at successive stages of ovarian development with abdomens containing four different quantities of fat body 145 Fig. 15. Proportions of host-seeking females of obligately autogenous species (i.e. Hybomitra fronta7is, H. hear7ei, and H. pechumani) at successive stages of ovarian development with abdomens containing four different ~uantities of fat body. 146
Chapter 3. Fig. 1. Trap head modified for use with data logger by addition of touch-sensitive electronic pad 166 Fig. 2. Mean, minimum, and maximum values of relative humidity, temperature, incident solar radiation, and number of tabanids • collected, recorded at half-hour intervals, 0530-2130 hr EDT, July 20 - August 5, 1991, at Iron Arm fen 167 Fig. 3. Number of half-hour intervals at temperatures ranging from 2 to 27°C, during which tabanids were not collected, and during which tabanids were active, as indicated by canopy trap captures; and percentage of intervals trapped during which tabanids were active, pl.otted against temperature 168 Fig. 4. Number of tabanids trapped per day at Iron Arm fen, and daily maximum temperature recorded at Schefferville airport, versus date, June-August, 1990 and 1991 ..•...... 169 Fig. 5. Temperature versus incident solar radiation recorded at each of 236 half-hour intervals during which tabanids were collected, Iron Arm fen, 1991 ...... •...... •.. 170 Fig. 6. Numbers of tabanids collected, June - August 1991 at Iron Arm fen, and period during which present study took place .... 171 • xvii • ACKNOWLEGEMENTS 1 take this opportunity to thank aIl of those who helped me throughout my
Ph.D. program. In particular 1 thank my supervisor, Dr. DJ. Lewis, for bis
support and guidance. 1 aIso extend thanks to the other members of my
examining committee, especially DIs. J.E. Burger and M.E. Rau, for their helpful
advice and critical review of this manuscript. 1 am greatly indebted to Amy
Chabot and Karen CIabb who, as summer students, slogged through bogs,
endured mosquito swarms and stressed-out graduate students, and generally slaved
on this projeet for many hours in the field and laboratory. 1 also thank Drs. HJ.
Teskey, Biosystematics Research Centre, and L L Pechuman, Comell University,
each ofwhom willingly shared with me bis extensive knowledge of the Tabanidae.
• 1 thank Pierre Langlois, ChiefTechnician, Entomology Department, McGill
University, for bis help, ideas, and peculiar humour, and Adla Halwani for her
help with the translation of the Abstraet. 1 also thank aIl those in the McGill
Entomology Department and at the McGill Subarctic Research Station,
Schefferville, who helped me out in many different ways. Very special thanks are
extended to staff and inhabitants of Franconia Farm, Howe Island, Ontario, for
keeping me sane. Last but not least, 1 sincerely thank my parents for aIl their
love, encouragement, and support, without which 1 would have been unable te
complete this projeet.
Initial funding for this projeet was provided by an NSERC operating grant • te DJ. Lewis. Further financial support was provided by the Department of xviii Entomology, McGill University, who awarded me Margaret M. Duporte • Fellowships in 1990 and 1991. Logistic expenses for the field work in Schefferville were largely covered by Northem Scientific Training Grants from the Deparnnent
ofIndian and Northem Affairs, through the Centre for Northem Studies, McGill
University.
•
• xix • FOREWORD While condueting the research which led to my Master's degree at the
University of Manitoba, 1 read a 1951 paper by L.A Miller which dealt with the
biology of hOISe flies and deer flies in northem Manitoba. Miller's work
addressed a wide variety of aspects of tile biology oftabanids in subarctic Canada,
but left a great many questions unanswered, particularly conceming the biology of
the 1arval stages. Subsequent reading made it apparent to me how little is known
ofthe biology of tabanids in the Canadian north, despite the fact that these
insects are notorious as blood-suckeIS throughout subaretic and boreal regions.
As an aside 10 my M.Sc. work in Manitoba, 1 was fortunate 10 be able to
conduct a briefstudy of the influence oftemperature and light intensity on
• tabanid host-seeking activity at Churchill, where Miller had carried out bis
original studies. This was my first visit to the north, and while there 1 caught the
northem 'bug' (literally as weIl as figuratively). When the opportunity presented
itself 10 conduct further research into the biology ofnorthem tabanids at McGill,
1 jumped. My objective was 10 cany out a detailed study of tabanid bionomics in
northem Quebec; this thesis is the result.
This thesis is comprised ofthree separate chapters, each dealing with a
different aspect ofthe tabanid biology in the vicinity ofSchefferville. Quebec.
Chapter 1 deals with the immature stages, Chapter 2, the adult stages, and
Chapter 3, the influence of weather on tabanid host-seeking activity. Since the • subject matter ofeach chapter diffeIS widely from that ofthe other chapters, and xx since each chapter deals with a separate body ofliterature. each chapter contains • a separate review of the pertinent literature. and a separate discussion. The combined chapters will, 1 hope. provide for a better understanding of the biology
ofTabanidae in subarctic areas of Canada.
•
• • 1
CHAPTERI
Distribution and Development of Larval Tabanidae (Diptera)
• in a Subarctic Peatland.
• 2 • J. LITERATURE REVIEW 1. BIOLOGY OF JMMATURE TABANIDAE.
The horse flies and deer flies, of the family Tabanidae, enjoy widespread
notoriety through their propensity for feeding on vertebrate blood, their
sometimes staggering local abundance, and their worldwide distribution. Ûoly the
blood-feeding female Tabanidae are well-known; few people are ever aware of
the existence of the non-blood-feeding males, or the immature stages of these
insects. Worldwide there are more than 4227 described tabanid species in 137
genera (J.E. Burger, unpublished data, 1992). Over 90% of these are known from
adults alone, with no information whatsoever about the immature stages (Burger
1977). Most studies of tabanid immature stages have focused almost entirely on • the rearing of final instar larva through to adulthood for taxonomie study. Consequently, for many species, only the habitat from which the larvae were
collected, what captive larvae fed on, and the duration of the pupal period are
known.
Most of what is known about the behaviour, ecology, and general biology
of immature Tabanidae has been derived from either (a) studies carried out under
relatively unnaturallaboratory conditions, (b) from the rearing of a very few
species from egg to adult, (c) from chance observation of egg, larva!, and pupal
stages in the field or (d) from systematic collection of larvae and pupae from the
field. From these varied studies a pieture has emerged of the generalized tabanid • life history, and sorne biological data (e.g., habitat choice, diet, etc.) which seem to hold true for most of those tabanid species for which larvae are known. lt has • often been assumed that what holds true for immatures of one tabanid species applies to others, and that what happens in one region is the same as wh:\t occurs
elsewhere. If, however, the biological diversity of tabanid larvae is anywhere near
that of better-known groups of biting flies such as mosquitoes, it is evident that
such assumptions may not be wise, and that we still have a very great deal to
learn about immature tabanids.
1.1. Lire Cycle.
1.1.1. Egg Stage.
Description. Typical of the Diptera, the life cycle of Tabanidae has egg, larval,
pupal and adult stages. Tabanid eggs are laid in small compact masses which • typically resemble drops of black or brown taro They may be single-layered, as are the masses of many Chrysops spp., or multi-tiered and conical, as are the
masses of most Tabanus and Hybomitra spp. (Stone 1930, Roberts 1966,
Pechuman 1981). When first laid, the eggs are white in colour, but quickly darken
as the chorion tans upon exposure to air. After a few minutes the outermost layer
of the mass typically assumes a black or dark brown colour (Pechuman et al 1983,
Graham and Stoffolano 1983). Individual eggs measure 1-3 mm in length (Hill
1921, Stone 1930).
Oviposition Sites. In most tabanid species for which oviposition habits are • known, the eggs are laid on emergent vegetation, leaves, branches, or similar such 4
objects overhanging water or moist soil (Stone 1930), a1though a few species • oviposit over drier substrates (J.E. Burger, pers. comm). Typicallocaùons include caltail and sedge blades at the edge of marshes or along ditches, and tree
branches or leaves overhanging ponds (Tidwell and Hays 1971, Pechuman 1981).
Exact choice of oviposition site is dependent upon species; a female may oviposit
on any available <;ubstrate located above potenùal Iarval habitat (Hill 1921,
Knudsen and Rees 1968), or her choice of egg-Iaying site may be very specifie.
Eggs of Tabanus stygius Say ar~ nearly always laid on leaves of littoral plants
(Saggitaria spp.) at the point where leaf meets stem (Hine 1906, Philip 1931), and
females of T. fairchildi Stone and T. làngi Austen only oviposit on rocks jutting
above the surface of rivers and streams (Hine 1906, King 1910b). Females of T.
abactor Philip are apparently unusual in that they oviposit directly upon the • substrate (Schomberg and Howell 1955).
Egg Mass Size. Depending on availability of oviposition substrate, and on
whether or not a female is disturbed during oviposition, she lays from one to four
egg batches per gonotrophic cycle (Stone 1930, Terterian 1983); the total number
of eggs laid varies considerably according to species. Large Tabanus spp. may lay
upwards of 500 eggs (Hine 1906, Hill 1921, Dukes and Hays 1971, Terterian
1983), whereas medium to small Tabanus, Hybomitra and Chrysops spp. generally
lay 200-300 (Isaac 1924, Schomberg and Howell 1955, Knudsen and Rees 1968,
Orimaù and Hansens 1974, Magnarelli and Anderson 1979, Magnarelli et aI. • 1980); the number of eggs laid per batch may be as low as 15 (Magnarelli and 5 Anderson 1979) or as great as 700 (Hill 1921). In addition to interspecific • differences in fecundity, intraspecific clif:erences are apparent: larger fe:n:!les invariably have more ovarioles than smaller individuals (Thomas 1971, Leprince
and Jolicoeur 1986, Leprince and Bigras-Poulin 1988), and therefore have the
ability to mature more eggs.
Development. Embryogenesis does not begin until eggs are laid (Hine 1906), and
usually requires 5-7 days (Hine 1906, King 1910a,b, Isaac 1924, Stone 1930, Segal
1936, Jones 1953, Orimati and Hansens 1974, Magnarelli and Anderson 1979),
although it may be accomplished in as little as 2 days (Knudsen and Rees 1968)
or as long as 10 (Hine 1906, Terterian 1983). Intraspecific variation in the time
required for embryogenesis under natura! conditions is related to temperature, • either as a consequence of eggs being laid in shaded vs. exposed areas (Isaac 1924), weather during incubation being cloudy vs. sunny (Stone 1930), or because
of latitudinal temperature variation (Terterian 1983). In the laboratory,
temperatures of 20-30' C are generally best for tabanid embryogenesis, with sorne
interspecific variation in temperature optima (Thompson et al 1979, Roberts
1980).
Hatching. AlI eggs in a single mass hatch more or less synchronously (Hine
1906, Hill 1921, Philip 1931, Segal 1936, Pechuman et al 1961, Orimati and
Hansens 1974) although in large, multi-tiered masses, successive layers of eggs • hatch over 1 or 2 hours (Terterian 1983). The relatively simultaneous hatching of 6 ail of the eggs in a mass is thought to be triggered by the mecbanical stimulus • caused when the first egg hatches (Stone 1930, Thompson et al. 1979, LL Pechuman, pers. comm). Hatcb usually occurs in the early morning, after eggs
have been warmed by the sun (philip 1931, ChvaIa et al. 1972, Teskey 1990).
1.1.2. Larval Stages.
Description. Tabanid larvae are a rather diverse group in terms of body form,
but the larvae of most North American species have a cbaracteristic elongate,
cylindrical, fusiform shape and usually white, yellowish, or various shades of green
or brown in colour. There are three thoracic and eight abdominal segments; on
ail but the terminal abdominal segment are three or four pairs of short prolegs
which are used in locomotion (Stone 1930). These prolegs are variously modified, • depending on the type of preferred habitat inhabited by the larva (Andreeva 1989). There is usually a pattern of dark pubescence on the thoracic and
abdominal segments, which can be of considerable value in separating the larvae
of different species (Teskey 1969). This pattern is most pronounced on late instar
specimens, and may not be visible on very small individuals (Tashiro and
Schwardt 1951). On the terminal abdominal segment there is a protrusile
respiratory siphon, and a single anal proleg. At the larva's anterior end is a
sclerotized head capsule. The head capsule is elongate and anteriorly-tapered,
with a prominent pair of slender, pointed, usually curved manchbles at the
anterior end, and a cephalic brush of spines located above the base of eacb • mandible (Burger 1977). The head capsule can be withdrawn into the first 7 thoracic segment, and this is often the case with preserved specimens (Teskey • 1969).
Development. Newly emerged tabanid larvae possess a sclerotized labral
hatching spine, which was used to split the chorion of the egg and allow the larva
to escape. The first Iarval instar apparently lasts only for a few seconds, since a
larva moults almost immediately after emergence (Hill 1921, Isaac 1924, Philip
1931, Roberts and Dicke 1964).
Second instar larvae are 1-3 mm in length, depending upon species, and
differ in appearance from first instars only in lacking the labral egg-burster (Bine
1906, Hill 1921, Philip 1931). They are relatively sluggish and do not feed,
apparently subsisting on yolk carried over from the egg stage (Bine 1906, Hill • 1921, Philip 1931, Chvala et al. 1972, Tenerian 1983). After 3-10 days, a larva moults to the third instar and, with newly scerotized mouthparts, begins to actively
seek food (Hill 1921, Isaac 1924, Philip 1931, Tenerian 1983).
Anywhere from 5 to 11 larval instars have been recorded in Tabanidae,
depending upon species and latitude. In Europe and what was formerly the
Soviet Union, larvae of the larger tabanid genera, e.g., Tabanus and Hybomitra,
pass through 7-11 instars, wher~ the smaller Chrysops spp. generally require only
6-7 (Chvala et al. 1972, Tenetian 1985, Olsujev 1977). In India, Tabanus rubidus
Wiedemann and T. striatus Fabricius larvae pass through seven instars (Isaac
1924), whereas larvae of Minnesota H. lasiophthalma (Macquart) pass through • eight instars (philip 1931). 8 When larvae of the Australian horse f1y T. townsvilii Ricardo were reared • from eggs, the majority of 14 day old larvae were 7-9 mm in length, but sorne had only reached 4 mm (Hill 1921). Similarly, 25 day old H. lasiop/uhaIma larvae can
vary from 3 to 7 mm in length (Hine 1906), and 10 day old T. atratus Fabricius
Iarvae vary from 3 to 6 mm in length (Stone 1930). Typically, a tremendous
amount of variation is observed in the developmental rates of conspecific tabanid
larvae under field and laboratory conditions, (philip 1931, Bailey 1948, Tashiro
and Schwardt 1953, Orimati and Hansens 1974), although larvae of certain
C/rrysops spp. may develop relatively synchronously (Burger 1977). Orimati and
Hansens (1974) noted that frequency of feeding and amount of food ingested per
feeding were the most important factors determining the rate of growth of captive
larval T. nigrovittatus Macquart, but Burger (1977) concluded that larval • development rate is also strongly affected by temperature.
Development Period. Most temperate tabanid $pecies are thought to have one
generation per year (Table 1), but som.e $pecies which typically develop from egg
adult in 9-10 months will facultatively add another year to their life cycles. In
these species, sorne individuals will remain as larvae until their third summer if
they are not large enough to pupate at the beginning of their second summer
(Stone 1930, Tashiro and Schwardt 1953, Magnarelli and Anderson 1978).
Sorne large tabanid species, or species in areas with a short growing season,
apparently always require two years or more years to complete their larval • development under field conditions. Conversely, rapidly-growing tabanid $pecies 9 may have 2 or more generations each year in areas with long grov:ing seasons • (Table 1).
Overwintering. Tabanids always overwinter as larvae. In temperate areas, the
onset of winter is accompanied by a graduai drop in substrate temperature; as the
substrate cools, tabanid larvae enter a quiescent, non-feeding diapause state, the
onset of which is triggered by photoperiod, temperature or a combination of the
two (Terterian 1983). In this state, tabanid larvae pass the cold season, often
trapped within a mass of frozen substrate (Hine 1906, Khan 1951, Terterian 1983,
Thompson et ai. 1979, Burger 1977, Roberts and Dicke 1964, Meany et aI. 1975).
Larvae do not migrate downward in the substrate to escape freezing (Schomberg
1952), as was once thought (Hine 1906). Larval diapause sirnilar to that observed • in temperate tabanid species occurs in tabanids inhabiting areas with a dry season. Annual aestivation enables tabanid larvae to pass the dry months in the drier
regions of India (Isaac 1924, 1925), Australia (Hill 1921). Africa (King 1910a,b).
and the southern United States (Burger 1977).
With the retum of conditions favourable to Iarval growth (i.e., spring thaw,
onset of rainy season), tabanid larvae break diapause and continue their
development. Ifa larva is of a species with a one-year life cycle, it feeds for only
a short time during its second season, and then pupates (Philip 1931). In at least
some species. a cold-conditioning period is necessary in order for pupation to • occur (Burger 1977, A Maire pers. comm.). Larvae of species with multiple-year 10 Iife cycles continue alternating growth and diapause states until their development • is complete.
Prepupal Stage. At the conclusion of its final instar, a tabanid larva ceases
feeding and, if it is an aquatic or semiaquatic species, migrates to a situation
where it is unlikely to be submerged in water for an extended period of rime (Hill
1921, Isaac 1924, Segal 1936, Roth and Lindquist 1948, Balley 1948, Thompson
1970, Ellis and Hays 1973, Teskey 1990). There it retraets its head capsule,
contracts its body, and undergoes a 1-7 day prepupal period (Stone 1930, Roberts
and Dicke 1964, Orimati and Hansens 1974), after which it positions itself
vertically just beneath the substrate surface and pupates.
• 1.1.3. Pupal Stage. Description. Tabanid pupae are obteetate and brown or straw-coloured, with a
stiff row of spines encircling the apical third of each abdominal segment (Teskey
1969, Burger 1977). At the apex of the abdomen are 6 stout, pointed projections
which form the aster, a structure which is apparently used by the pupa to change
its location within the pupal burrow (Segal 1936, King 1910b, Philip 1931,
Schomberg 1952).
Duration of Pupal Period. The pupal stage usually lasts approximately two
weeks, but may vary considerably among species, even at the same location (Table • 2).. The pupation period of Chrysops spp. is usually shorter than that ofHybomitra 11 spp. at the same latitude. At the conclusion of pupation, the cuticle of the pupal • prothorax splits and the adult tabanid emerges (Roth and Lindquist 1948).
1.2. Larval Diet.
The larvae of practically all known tabanid species are considered to be
predacious (Teskey 1969, Burger 1977) The ooly exceptions are larval Cluysops
spp., which are widely considered to be detritivorous (e.g. Pechuman et al 1983),
based mainly upon speculative evidence. It is more likely that Cluysops larvae are
in fact predaceous upon very small soil invertebrates, owing to the morphological
similarity of their mouthparts of more predatory tabanid larvae (J.E. Burger, pers.
comm.). Carnivorous tabanid larvae will apparently attack "any soft-bodied
organism unfortunate to cross their paths" (Isaac 1925), although species exhibit • markedly differing degrees of aggression (King 1910a, Isaac 1924, 1925). Recorded prey items include annelids, molluscs, crustaceans, and other inseets
(Hine 1906, Jones 1953, Gingrich and Hoffman 1967, Hine 1906, Meany et al.
1976), and occasionally small amphibians (Jackman et al. 1983). Cannibalism
among tabanid larvae is thought to be a frequent occurrence (Isaac 1924, Philip
1931), and is considered to be a major factor limiting larval populations of at least
some species (King 1910a, Gingrich and Hoffman 1967, Meanyet al. 1976). In
captivity, the tendency of larvae caged together to eat one another has proven a
major barrier to mass colonization, since the larvae must be reared separately to
prevent them from devouring one another. Not all tabanid species are • cannibalistic, however, and some species (e.g., Tabanus par Walker, T. abactor 12 Philip, Hybomitra lasiophthalma) can be mass-reared without much difficulty, • (King 1910a, Schomberg and Howell 1955, Thompson et al. 1980). In Japan, the larvae of severa! Tabanus spp. are common inhabitants of
rice paddies, where it is customary for labourers to work without shoes or gloves.
Not uncommonly, labourers are bitten by the tabanid larvae; the bite causes a
burning pain and considerable swelling to the vietim, as a consequence of
injection of larva! venom into the wound site (Otsuru and Ogawa 1959).
1.3. Larval Habitats.
Terterian (1985) and Andreeva (1989) categorized the larval habitats of
many tabanid species which occur in the former USSR, and placed species in
distinct groups according to larval habitat preference; these groupings are also • applicable to the habitat preferences of larvae of nearctic tabanid species. a) Semiaquatic (= Hemibydrobionts). Larvae of the majority of nearctic
Hybomitra and Atylotus spp. and many nearctic Chrysops spp. typically inhabit
semiaquatic habitats; e.g., moist soil or moss at the margins of lakes or ponds,
moist soil in low-lying, poody drained, or seepage areas, or moss in peatland
habitats (Teskey 1969, Lane 1976, Burger 1977).
b) Aquatie. Lentie (= Hydrobionts: Limnophils). Larvae of some species are
aquatic in the strict sense of the word, inhabiting the bottom muck in pond or
lake habitats for mos! of their larval development, and only migrating above the • shoreline immediately prior to pupation. Larvae of many nearctic Chrysops and 13 several Hybomitra spp. fall into this category (Teskey 1969, Lane 1976, Burger • 1977). c) Aquatic - Lotic (= Hydrobionts: Rheophils and Subrheophils). A rather
small and specialized aquatic group includes larvae which dwell under rocks in
flowing-water habitats. The larvae of species in this group often possess elongate
prolegs or crochets as an adaptation to avoid being swept away by the current
(King 1910b). Nearctic representatives of this group include Tabanus faircJùIdi, T.
domfer Walker, and T. abditus Philip (Burger 1977, Pechuman et aL 1983).
c) Terrestrial (= Edaphobionts). Larvae of a few tabanid species are adapted to
living in relatively dry soils, such as might be found in pastures or beneath the leaf
liner on forest floors. Two common horse fly species, T. quinquevittatus
Wiedemann and H. lasiophthalma, are local examples of terrestrial tabanid • species (Tashiro and Schwardt 1949, Logothetis and Schwardt 1948). The aforementioned groupings are by no means absolute; there are species
which are transitional between each group, and also habitat generalists or
specialists within each group. For exarnple, the larvae of T. marginalis Fabricius
and T. similis Macquart have been recorded from a wide range of semi-aquatic .
habitats (Teskey 1969, Thompson 1971), whereas larvae of H. minuscula (Hine)
oCcu! only in sphagnum bogs (T~skey 1969, Pechuman 1981). The habitat of
larval tabanids is largely determined by a female fly's choice of oviposition site,
since the larvae are thought to usually remain within a hundred metres of where
their egg mass was laid (Lane 1976). Under most circumstances, tabariid larvae • only inhabit the top 10 cm of substratt: (Segal 1936, Bailey 1948, Khan 1951, Wall 14
and Jamnback 1957, Ellis and Hays 1973), but can occur at greater depths, • depending on substrate and circumstances.
1.4. Larval Densities.
Densities of larval tabanids in different habitats vary enormously. Very
2 high densities of aquatic species, in some cases over 1400 larvae/m , may be
encountered when late instar larvae congregate along the margins of ponds or
lakes prior to pupation (Roth and Lindquist 1948), although larval densities of 14
26/m2 are more commonly observed (Knudsen and Rees 1968, Lane 1976,
Burger et ai. 1981). In contras!, earlier instars are seldom colleeted at densities
greater than 10 larvae/m2 (Ellis and Hays 1973, Lane 1976, Meany et ai. 1976).
Among semiaquatic species, larvae of salt marsh tabanids are the best studied, • owing to the high densities at which they are often encountered. Larval densities 2 in salt marshes average from 8 to 48/m , varying greaùy within and among
marshes depending on vegetation, soil type, degree of tidal flooding, and general
soil moisture regime (Bailey 1948, W~ and Jamnback 1957, Freeman and
Hansens 1972, Dukes et ai. 1974). In other types of semiaquatic habitats, larval
densities vary widely, since larvae are apparenùy distributed very unevenly
throughout available habitat. High larval densities may, however, be encountered
in pockets of optimal habitat (Ellis and Hays 1973, Lane 1976). Lane (1976)
sampled six types of tabanid larval habitats in California, and found that average
densities varied from 14-15 larvae/m2 in seepage areas and temporary pond • margins, to <7Iarvae/m2 in other habitats. High densities of larvae have aIso 15 been reported from wooded swamps (Thompson 1970), banks of permanent ponds • (Lewis and Jones 1955), pasture streams (Tashiro and Scbwardt 1949). and margins of ponds containing dairy barn effluent (Gingricb and Hoffman 1967). In
terrestriaI habitats, densities of 4-21 larvae/m2 have been recorded in pasture soil
(Logothetis and Scbwardt 1948, Schomberg and Howell 1955, Tashiro and
Scbwardt 1953), and 10 larvae/m2 in forest soil (Wilson 1969).
Within a habitat type, tabanid larvaI density is affected by factors sucb as
substrate composition and compaction (Ellis and Hays 1973), soil moisture level
(Khan 1951, Tashiro and Schwardt 1953), and availability of suitable oviposition
substrate for adult femaIes (Knudsen and Rees 1968). Time of year when
sampling occurs is aIso important in determining numbers of larvae colleeted; in
areas where most species have one-year life cycles, sampling carried out in the • spring will colleet mostly large, easily seen larvae, whereas only smaIl, easily missed larvae will be taken if sampling is carried out in autumn (Lane 1976).
Because numbers of tabanid larvae vary so greatly between and within habitats,
care must be taken when interpreting estimates of the number of tabanid larvae
present in an area, or estimates of an area's potentiaI adult production which have
been extrapolated from a limited sampling of larvae (e.g., Tashiro and Schwardt
1949, Miller 1951, Gingrich and Hoffman 1967, Ellis and Hays 1973).
1.5. Natural Enemies of Immature Tabanidae.
Egg Stage. Tabanid eggs are parasitized by severa! species of minute • Hymenoptera, in the families Trichogrammatidae, Scelionidae and Mymaridae 16 (Dukes and Hays 1971, Jackson and Wilson 1965). Parasitism rates of 40 to 90% • have been reported (Segal 1936, Jones 1953, Dukes and Hays 1971, James 1963), indicating that these parasites may be important in regulating tabanid populations
in certain situations. In addition to parasites, larval and adult coccinellid beetles
of severa! species have been observed to voraciously consume tabanid egg masses
(Jackson and Wilson 1965, Magnarelli et al. 1980, Orimati and Hansens 1974).
Larval and Pupal Stages. Larval tabanids are parasitized by Diptera
(Tachinidae, Bombyliidae), Hymenoptera (Pteromalidae, Diapriidae) (Magriarelli
and Anderson 1980, James 1963, Jones 1953, Segal 1936), mermithid nematodes
(James 1963), as weIl as fungi and microsporidia (J.E. Burger,per.r. comm.). The
majority of larval parasites mature and emerge during the pupal stage of the host, • and may thus be considered pupal parasites as weIl (Teskey 1969). In genera! larval and pupal parasitism is not thought to occur at levels high enough to impact
tabanid host populations (James 1963).
Among predators of tabanid larvae and pupae, fish (Hine 1906) and
shorebirds (Meany et al. 1976) are thought to take numerous tabanid larvae and
pupae, particularly in coastal salt marshes wbere high densities of tabanid larvae
and sborebirds occur. Sperry (1940) and Booth (1968) found tabanid larvae in the
stomacbs of 7-11% of common snipe Capella gallinago (Linneus) collected in
Louisiana rice fields, and Tuck:(1972) reported that Hybomitra and Chrysops
larvae comprise up to 14-30% of the diet of snipe in peatlands in northem • Manitoba and Newfoundland. 17 The most imponant predators of larval tabanids, however, may very wel1 • be their own cannibalistic kin; the cannibalistic habits of horsefly larvae are suspeeted to have a greater effect on larval survival than do all predators and
parasites combined (Pechuman et ai. 1983, Segal 1936).
2. IMMATURE TABANIDAE IN SUBARcnC CANADA.
The subarctic region extends in a broad belt between the forested boreal
region and the treeless aretic, and covers approximately 760,000 km2 of nonhem
Canada (National Wetlands Working Group 1988). During the brief summer in
the subaretic, tabanids, particularly Hybomitra spp., are extremely common, and
are legendary for their tenacious barrassment of humans and animals. As
elsewbere, praetically nothing is known of tabanid biology or life history in • nonhem regions, although the immature stages have been described for 76% (32/42) of tabanid species which oceur in the Canadian subarctic (Teskey 1969,
1984, 1990). This discrepancy stems from the faet that most species which are
found in the subaretic are aIso found in southem regions of the country, where
larval collection and rearing bave been carried out.
The abundance of tabanids in the subaretic is not surprising. Larval
habitat is COmmon in the region, where wetlands constitute about 30% of the land
surface (National Wetlands Working Group 1988). Adult flight seasons of
tabanid species in the subaretic summer are often sboner and more prone to
overlap those of other species than the fligbt seasons of tabanids in southem • Canada (BanDeau and Maire 1983b, Twinn et al 1948, Miller 1951, McElligott 1S and Galloway 1991), resulting in a much greater apparent total abundance of • adults on any given day during the summer. FinaIly, large cervids such as woodland caribou (Rangifer tarandus Gmelin) and moose (Alces alces Linnaeus) in
the subarctic provide a relatively common source of blood for host-seeking
tabanids.
The work of Miller (1951) at Churchill, Manitoba (5S046'N; 96°11'W) has
provided us with much of what is known of the biology of immature tabanids in
northem Canada.
Eggs. The oviposition sites of tabanid females in subarctic peaùands, or any
peatlands for that matter, are very poorly known. Miller (1951) reported not
finding egg masses on vegetation surrounding larval habitats, despite a thorough • search during the months of July and August, however Shemanchuk (in Miller 1951) and James (1952) found large Hybomitra spp. egg masses on dwarf birch
(Betula glandulosa Michx.) overhanging pools. Since eggs are laid on emergent
vegetation by nearly aIl of those Hybomitra" and Chrysops spp. for which
oviposition habits are known, it seems safe to assume that the eggs of species in
the subarctic are also laid on leaves, branches, and stems.
Larvae. Miller (1951) sampled a variety of weùand habitats in the vicinity of
Churchill, Manitoba, and found the highest concentrations of larvae in organic
matter on the bottom of tundra ponds, and in the top '10 cm of substrate in wet .: tundra meadows and muskeg. James (1952) also colleeted Hybomitra and 19 CJ:rysops larvae from a bog and tundra meadows near Churchill. Miller estimated • larval densities to be from 17 to 50 larvae/m~ (67,200-200,900 larvae/acre) in the most productive larval habitats, but noted that many habitats contained few or no
larvae. An important consideration of Miller's (1951) study is that larvae were
collected by hand-sorting, suggesting that very small tabanid larvae (e.g.• 2·5 mm)
wcre aImost invariably missed; therefore his estimates of larval abundance are
almost certainly low. Miller's larval density estimates resemble those obtained by
Lutta (1970) and Lutta and Bykova (1982) in taiga regions of Karelia, in what was
formerly the northem USSR; they found that the larval stages of tabanids were
distributed very unevenly within semiaquatic habitats. and that while densities
2 2 ranged from 5-30 larvae/m • they averaged 5-6/m owing to the diffuse area-wide
distribution of the larvae. In subarctic Siberia, Olsufev (1949) and zapekina • Du'keit (1966. 1969) reported up to 48 larvae and pupae/m2 in sedge hummocks of peatlands. In southem Canada, Baribeau and Maire (1983a) also found that
tabanid larval densities varied considerably among peatlands and arnong sampling
2 locations in peatlands. but averaged 15 larvae/m •
Miller estimated that the life cycles of Hybomùra and C/uysops spp. in the
Churchill area were at least 3 years, based upon observations condueted under
semi-natural conditions. Of relatively large (i.e.. >1 year old) larvae he colleeted
during the summer of 1949 and maintained in an outdoor enclosure. many
emerged the following year. but about one half remained as larvae until at least • the spring of 1951, when the study was terminated. 20 Miller observed that larval Hybomitra spp. at Churchill were exclusively • predators, and fed willingly on the larval Tipulidae (Diptera) which aImost invariably shared their habitat. In addition, they were observed to feed on snails
and annelids. Captive Chfysops larvae, on the other hand, did not require feeding,
and apparently subsisted entirely on organic material present in their rearing
substrate.
Pupae. The pupae ofHybomitra spp. in the subarctic, as elsewhere, are found
just below the substrate surface. They are always found in drier habitats than
those inhabited by the larvae, in sorne cases up to 18 m (20 yd) from the larval
habitat (Miller 1951). James (1952) colleeted Hybomitra pupae from moss
hummocks in bogs in the Churchill area. Miller found that C/uysops pupae, on • the other hand, were usually found in the immediate vicinity of the larval habitat, in water-saturated material (e.g., at the edge of pools), and were invariably found
2 2 at densities of 32-129/m (3-12/ft ). Pupation periods of laboratory-reared
Hybomitra spp. averaged 22-27 days (range 18-28), whereas those of Chrysops spp.
averaged 17-18 days (range 15-21). Miller aIso noted that pupation occurred on
approximately the same dates in the laboratory and field, even though the
laboratory temperature was approximately SC C (9SO F) higher than that in the
field. James (1952) reared larvae of severa! tabanid species from the Churchill
area; pupation dates he reponed for each species agree closely with those
recorded by Miller, which suggest that the onset of pupation for larvae a given • species varies relatively little from year to year.
;, 21 Natural Enernies. The chalcidoid wasp Diglochis occidentalis (Ashmead) is the • most imponant parasite of tabanid immatures in the Churchill area, attacking 13% of Hybomitra larvae and 21% of C/uysops larvae (James 1952). The wasp
lays eggs on a large tabanid larva during the summer months, but its larvae do not
devour the tabanid larva until it pupates, usually the following spring (Miller
1951). Mermithid nematodes a1s0 parasitize tabanid larvae at Churchill (James
1952).
LarvaI C/uysops are preyed upon by larvae of the tipulid Prionocera
dimidiata Leow in wetlands near Churchill (Miller 1951, James 1952); these
tipulid larvae, in turn, are preyed upon by large Hybomitra larvae (Miller 1951).
Canmbalism by large Hybomitra larvae was considered by MilIer to be a major • factor regulating larvaI tabanid populations at Churchill.
• 22 • Table 1. Recorded lengths of tabanid Iife cycles in the laboratory and field.
Spccles Larval Pcrlod location Reference
Çh~ysops se1unctus Szilady 1 ye4r (l) USSR Terterlan 1983 c. v'tt~tus Wtedemann 2 gen'/year (l) Florlda Jones 1953 Cl vittatus Wtedemann 1 year (l) New York logothetls &Schwardt 1948 Hybomltra caucasie, (Enderletn) 2 year (l) USSR Terterian 1983 H. lastopht~lma (Macquart) 1 year (l) New York l090thetls &Schwardt 1948 H. laslophthalma (Macquart) 1 year (li Minnesota Philip 1931 H, ntttdtfpons (HcOunnaugh) 2 year (l) Wisconsin Roberts and Olcke 1964
Nemortus caucastcU$ [Olsufjev) 1 year (l) USSR Terterlan 1983
leucotabanus annulatus Say 2 year (F) Mississippi lewls &Jones 1955
Tab~nus abactor Philip 1 year (F) Oklahoma Schomberg &Howell 1955 T. atratus Fabrtcius -240 day (l) Florlda Jones 1953 T. autumnal'; Ltnneus 206-317 day (l) USSR Terterlan 1983 T. bromius Ltnneus 703-743 day (l) USSR Terterlan 1983 TI endymion Dsten Sacken 259-330 day (l) Florlda Jones 1953 T, fumiP$"nts Wiedemann 345 day (l) Florlda Jones 1953 T. lcleani Austen 241-295 day (l) USSR Terterian 1983 T. lineola Fabrtctus 2 gen./year (l) Florlda Jones 1953 T. lincola Fabrtctus 240-244 day (l) Massechusetts Khan 1951 T. ltneala Fabrtctus 2 year (F) New Jersey Frceman &Mansens 1972 • T. ntgrovtttatus Macquart 2 gen./year (l) Florlda Jones 1953 T. ntgrovtttatus Macquart 2 year (F) New Jersey Frceman &Mansens 1972 T. ntgrov'ttatus Macquart 2 year (F) Massechusetts Heany et al. 1975 T. ntgrovtttatus Macquart 1-2 year (l,F) Connecticut Magnare11 1 and Anderson 1978 T. ntgrovtttatus Macquart 1 year (l) Texas Thompson 1979 T. gutnguevittatus Wiedemann 1-2 year (l.F) New York logothetis &Schwardt 1948 ~ Walker 96-149 day (l) Sudan King 1910a T. rufiventris rabricius 3 gen'/year (l) Indla Isaac 1924 T. similis Macquart 5 month (l) Wisconsin Roberts and Olcke 1964 T. speetabilis Loew 1 -2 year (l) USSR Terterian 1983 T. striatus Fabricius 3 gen./year (l) lndla Isaac 1925 T. stvaius Say 1-2 year Minnesota Philip 1931 T. stvaius Say > 1 year (l) U.S. Hine 1906 T. sulcifrons Macquart 1-2 year (F) Oklahoma Schomberg 1952 T. townsvilli Ricarda - 190 d (l) Australla Hill 1921
Varlous spp. 1 or 2 year Ontario Pechuman et al. 1961 Various spp. 1 or 2 year (l.F) Arizona 8urger 1977 Vartous spp. 1 or 2 year (F) Alabama E111 s &Mays 1973 Vartous spp. 1 or 2 yr (F) New York Tashlro &Schwardt 1949 Various spp. 1 ta 6 %yr (l.F) USSR Terterlan 1949
l • laboratory rearlng study /'-': . • . ~- F• Field observations of larval slze distribution • Table 2. Recorded lengths of tabanid pupaùon periods in the laboratory.
Spccles Pupatlon Perlod location Reference (days)
Chrysops aberran; Philip 7 Wisconsin Roberts and Dleke 1964 C. carbonarius Walker 8 New York Stone 1930 C. aestuans van der Wulp 11 - 21 (avg. 11) $W Ontario James 1963 C. discalls WIlllston - 2S Oregon Roth &lindqulst 1948 C. excita"s Walker 8 - 9 Minnesota Phl1lp 1931 C. mltis Dsten Sacken 8 - 9 Minnesota Phl1lp 1931 C. pikei Whitney 4 Wisconsin Roberts and Dlcke 1964 C. vlttatus Wtedemann 13 New York Stone 1930 C. vittatu$ \liedemann Il - 21 (avg. Il) sw Ontario James 1963 Hybomttra lastophthalrna (Macquart) 15 U.S. Hine 1906 H. lasiophthalma (Macquart) 11 - 15 New York logothetls &Schwardt 1948 H. lasiophthalma (Macquart) 13 - 16 Minnesota Phl1lp 1931 H. lurlda (Fallen) 9 Wisconsin Roberts and Dlcke 1964 H. nltldlfra"s nuda (HcDunnough) 10 Wisconsin Roberts and Dick. 1964 • H. sodalls (Wllliston) 17 sw Ontario James 1963 Tabanus atratU$ Fabrtcius 14 New York Stone 1930 T. abactor Philip 14 Oklahoma Schomberg &Howell 1955 T. calens linnaeus 10 New York Stone 1930 T. crassus \la1ker 17 - 19 lndla Isaac 1924 T. fumipennts Wiedemann 9 - 11 Florlda Jones 1973 L..o!!: Walker 6 - 8 Sudan King 19100 T. reinwardtii Wtedemaron 17 New York Stone 1930 T. rubidus Wiedemann 7 - 8 l"dla l..ac 1924 T. striatus Fabrictus 5 lndla Isaac 1925 T. townsvi'" Ricardo 8 - 20 (avs. 12) Australla Hill 1921 Atvlotus aqrestts \liedemann 6 Sudan King 191Db Chrvsops, Nemorius spp. 3 - 18 US5R Terterlan 1963 Tabanus Hybomitra spp. 10 - 33 USSR Terterlan IS63 Various spp. 14 - 21 Ontario Pechuman !l.Lll. 1961 Varieus spp. 5 - 14 New York Segal 1936 Varicus spp. 5 - 14 Florida Jones 1973 • 24 • II. INTRODUcrION Apart from the work of Miller (1951), no study has examined the biology
of the immature stages of Tabanidae in the Canadian subarctic, despite the faet
that these insects are common biting pests of mammals throughout the region.
This lack of knowlege represents a major void in our understanding of northem
biting rues, and also in our understanding of the peatland ecosystems where larval
tabanids live. Much of what has been recorded in the literature conceming the
biology of northem tabanids has been extrapolated from tabanid species living
either in the United States or other parts of the world.
1 earried out this study with the intent of attempting to fill some of the
gaps in the knowlege of immature Tabanidae in northem Canada. My goal was • to obtain information concerning the biology of larval tabanids from a peatland habitat near Schefferville, Quebec, to (a) determine the spatial distribution,
relative abundance, and habitat preferences of larval tabanids within the peatland
and (b) to determine the annual rate of larval development for tabanid species
common in the Schefferville area. To this end, 1 colleeted larval tabanids from
early June to late August in 1990 and 1991, and made observations on other
aspects of tabanid biology in the vicinity of the peatland. This study was
condueted in conjunetion with a study of the biology of adult horse flies and deer
flies near Schefferville, which constitutes the second and third chapters of this • Ph.D. dissertation. 25 • III. MATERIALS and METHODS 1. Study Area.
Larval Tabanidae were collected from Iron Arm fen, a 580 m x 110 m
minerotrophic peatland located approximately 20 km NE of Schefferville, Québec
(540 49' N, 660 W), near the Québec-Labrador border (Fig. 1). Eight sampling
sites were selected within the fen to represent as broad as possible a range of
moisture and vegetation characteristics (Fig. 2, Table 3). Four of these sites were
in relatively dry regions of the fen, where the water table was 3-5 cm below the
substrate surface. Two sites were located in wet regions of the fen, where the
water table was at or slightly below the substrate surface. Two other sites were in
very wet regions of the fen, where the substrate was submerged. • At each study site, a lx1 m wooden frame was randomly placed on the substrate; the percent cover of vegetation, exposed substrate, and open water
within the frame were estimated. Values for percent cover presented in Table 3
are based upon the average of three frames per site.
Temperature and pH were recorded at each site whenever substrate
samples were taken. Measurements were taken in the water which pooled in the
weIl which was left after a sample had been taken. Temperature was measured
using a: standard alcohol thermometer; pH was taken using a portable pH meter
(Corning™ Model PS-15), which was calibrated weekly. Mean pH values • obtained at each site are presented in Table 3. 26 pH was apparently quite unifonn throughout the fen, except that at Site 5 • it was significantly higher than at the other sites (ANOVA: df=7, p summer (ANOVA: df=l1, p>0.05). Temperature did not differ significantly among sites (ANOVA: df=7, p=0.5276), but varied according to sampling date (ANOVA: df=l1, p successive weeks of sampling 1990 and 1991 are presented in Fig. 3. 2. Sampling. In 1990 twice-weekly sampling at four sites (Sites 1-4, Fig. 2) was earried out for 11 weeks (14 June - 29 August), as follows: • 2.1. Wet Extraction. A 10 x 10 cm template of hardware cloth was placed on the substrate, and a serrated knife used to cut a 500 cm3 substrate sample (lOx10xS cm = length x width x depth). Two samples were taken from each site and each was placed in a separate, labelled plastic bag for transport to the laboratory. In the lab, samples were processed as follows: Each sample was placed in a 1 cm mesh hardware cloth basket (12x12x7 cm), which was then set in a plastic container (12x12x14 cm) filled to the top with tap water. Each container was then covered with a sheet of clear plastic to prevent evaporative moisture loss and escape of organisms. Containers were then placed in an bath of ice water 3 cm deep, 4 • containers per bath. One 100 watt bulb with an aluminum refleetor was 27 suspended at a height of 10 cm over every two samples (Fig. 4a). This • arrangment was maintained for 24 hrs, with ice being replenished as needed. At the conclusion of the e.xtraction process all peat samples were discarded, and larval tabanids and other invertebrates remaining in the water bath were removed and sorted. Fairchild et al. (1987) found that no further invertebrates could be extracted from Sphagnum samples by continuing wet extraction for periods longer than 24 hours. 2.2. Dry Extraction. Two dry extractors similar in design to that described by Teskey (1962) were construeted of pine two-by-fours. Each extraetor consisted of shallow, 100 cm x 200 cm reetangular frame partitioned into four equal 5Ox100xS cm compartments. • The bottom of each compartment was covered with 05 cm mesh galvanized screen. Undemeath the screen at the bottom of each compartment was mounted a downward-pointing funnel of black plastic sheeting. A hinged, aluminum foil lined lid 5 cm deep was constructed so that it could be lowered to coyer the top of the extraetor chambers. Eight eleetrical sockets with 60 watt bulbs were secured i::. J this lid, two above each chamber, such that the bulbs was suspended approximately 8 cm above the screen bottom of each chamber. Each funnel emptied into a shallow dish of water which served as a colleeting container (Fig. 4b). A single 4500 cm3 substrate sample (30x30xS cm) was cut from the peat at • each sampling site using a smal1, sharpened, flat-bladed spade. Samples were 28 carefully placed in labelled 20 1plastic buckets for transport to the laboratory. In • the lab, peat samples were placed in the dry extraetors, one sample per chamber. At the beginning of the extraction procedure, samples were cut into strips approximately 5x5x30 cm to facilitate drying, and the strips spread out on the wire racks. The !id was then closed, the !ights tumed on, and the samples left to dry. Tabanid larvae and other invertebrates were colleeted daily from each colleeting container in order to reduce loss due to the presence of predators in the container (e.g., Odonata naiads, other tabanid larvae). The extraction process continued for approximately 72 hours, by which time peat samples had dried completely. . In 1991 sampling was condueted at all eigbt sites (Sites 1-8, Fig. 2) for 12 weeks (7 June - 29 August) as described above, except that wet extraction samples • were colleeted once, rather than twice, weekly. In both 1990 and 1991, tabanid larvae and pupae in addition to those colleeted at the eigbt sampling sites were obtained sporadically from a varlety of sources e.g., moss hummocks, samples taken as part of another study. 2.3. Emergence Traps. Througbout the summer of 1991, four emergence trapS were installed over moss hummocks in Iron Arm fen. Bach trap consisted of black plastic sheeting stretched over a pyramidal wooden frame, and enclosed a lx1 m area of substrate. .At the apex of each trap, a removable, clear plastic colleeting container was • affixed ta permit collection of emergent inseets. 29 3. Preparation and Treatment of Larvae. • AlI tabanid larvae colleeted were either immediately killed in KAAD solution, which causes soft-bodied invertebrates to fully e.xtend (Martin 1977). and measured; or placed in an ice-water bath until fully extendec!, and then measured (Teskey 1969). Most measurements were made using a calibrated ocular micrometer inserted into the eyepiece ('If a stereoscopic microscope; e.xceptions were the total body lengths of large Hybomitra larvae, which were measured using a milljrneter rule. Total body lengths were measured from the anterior tip of the labium on the head capsule to the posterior-most tip of the extended respiratory siphon. Head capsule lengths were measured from the anterior tip of the labium to the posterior bilobed extremity of the epicranium. Most larvae were kept alive throughout the summer of 1990. Individual • larvae were maintained in capped 120 ml plastic urine specimen containers containing a small amount of moistened peat. These larvae were fed once weekly, either with a freshly-killed mosquito larva or pupa (Aedes spp.) for small Chrysops larvae, or a living Trichoptera larva (ümnephilidae) for larger Hybomitra larvae. It should be noted that the reputedly detritivorous Chrysops larvae willingly fed upon living mosquito larvae andf pupae. Larvae were measured weekly throughout the summer months, from their date of capture until they diec!, or until the end of August, when all still-living larvae were killed in KAAD. Killed larvae were preserved in 70 % ethanol. Species identifications of • larvae were made using Teskey's (1969, 1983) keys. 30 • RESULTS and DISCUSSION 1. Larval Diversity. Tabanid larvae were extraeted from a total of 'lJ37 peat samples (129 m3 total volume) by dry extraction, and from 360 samples (0.18 m3 total volume) by wet extraction. A total of 406 tabanid larvae were colleeted: 96% by dry extraction, 4% by wet extraction. Even when a correction is made for the faet that cach dry extraction sample was nine times greater in volume than each wet 3 3 extraction sample (4500 cm vs. 500 cm ), significantly more tabanid larvae per unit volume of substrate sampled were taken by dry extraction than by wet extraction (1.32:!:0.11Iarvaejlitre vs. 0.40:!:0.11Iarvae/litre; ANOVA; df=!, p bogs (Teskey 1984). Numbers of larvae collected of each species are listed in Table 4. 1.1. Chrysops spp. Of the 357 C/zrysops larvae collected, 85% were identified to species. The remaining 15% were not identified, either because they were too small (species diagnostic patterns of pubescence were not reliably visible on larvae >4 = in length), or because the larvae escaped, were misplaced, or were badly damaged • during collection. 31 Larvae of C ater Macquart, C. frigidus Osten Sacken, C furcatus Walker, • and C nigripes Zetterstedt were identified. Of the first three species, each comprised about 5% of the Cluysops larvae collected. Thirty percent of Cluysops larvae colleeted were identified as C nigripes; they differed slightly from individuals of this species by Teskey (1969)(Fig. Sb.c). Larvae of all the aforementioned species except C ater have been colleeted previously as larvae in peatland habitats (Miller 1951, Baribeau and Maire 1983a, Teskey 1969, Pecbuman 1981). and all were relatively common at the fen as adults. Forty one percent of the Chrysops larvae colleeted displayed a pubescence pattern whicb did not fit that of any described species (Fig. Sa). and are presumed to be the larvae of Cluysops zinzalus Philip. In both 1990 and 1991. C zinzalus was the most commonly collected species among the adult Chrysops at the site; • likewise the undescribed larvae were most abundant among the Chrysops larvae colleeted. Adults of C zinzalus and C nigripes are similar in appearance and are thought to be closely related (Teskey 1990); larvae of the undescribed species closely resemble larval C nigripes. Finally. a pupa, from which a C zinzalus Q emerged, was colleeted from a moss hummock at Iron Arm fen; this proves that larvae of C zinzalus were present in the fen. Adult C zinzalus are frequently colleeted in the vicinity of peatlands in northern and southern Quebec (BanDeau and Maire 1983b). • 1.2. HybomiJra spp. • Ooly 47 Hybomitra larvae were collected from Iron Arm fen. Nineteen of these (29%) were identified as being Hybomitra arpadi (Szilâdy), the species most common as adults at the site. Also colleeted were 5 specimens of H. pedlumani Teskey & Thomas, 2 specimens of H. zonaIis (Kirby), 2 H. epÏSlates (Osten Sacken), and a single H. itasca (Philip) (Table 4). Oddly, no adults of these latter two species were colleeted during the adult trapping of the current study, although they have been reported previously from the Schefferville area (Teskey 1990, I..L Pechuman pero. comm.). Forty percent (19) ofHybomitra larvae could not be identified to species, even though the majority were of identifiable size. This is not surprising, since larvae of four of the 14 Hybomitra spp. known to occur in the Schefferville area • are not yet described; larvae are unknown for H. aequetincta (Becker) and H. hearlei (Philip), two of the most common species in adult collections at Iron Arm fen, and H. astuta (Osten Sacken) and H. liorhina (Philip), two uncommon species. 2. Pupae. A determined search of moss hummocks in the fen in June 1990 yielded 12 intact tabanid pupae and 21 pupal exuvia. Nine H. arpadi pupae were colleeted between 27 June and 4 July. Adults emerged from three of these; on 2 July (1 ct), 3 July (1 ct), and 5 July (19). One unknown HyhomiJra pupa was colleeted on • 28 June, and a badly damaged Chrysops pupa was colleeted on 25 June; both diec!. 33 As was reported earlier in this section, a pupa was colleeted on 5 Juiy from which • a C zinzalus 9 emerged on 8 July. Exuvia, all relatively weil preserved, were collected from 3-10 July. Eighteen were H. arpadi exuvia, 2 were those of other Hybomitra spp., and one C 2 nigripes eX".Ivium was colleeted. In a single moss hu=ock approximately 1 m , 14 H. arpadi exuvia~ were colleeted. Late instar tabanid larvae have been reported to consregate at high densities prior to pupation in habitats which are drier than those in which they developed (Roth and Lindquist 1948, Miller 1951, Lane 1976). This aggregration presumably occurs when larvae migrate to higher ground to avoid high moisture levels and possible ~owning during the pupal stage; high densities are tolerated because larvae do not feed i=ediately prior to pupation (Thompson 1970, Ellis and Hays 1973). Miller (1951) reported that Hybomitra • larvae at Churchill migrated to drier habitats prior to pupation, but did not mention pupal density. zapekina-Du'keit (1966, 1969) and Olsurev (1949), however, found that tabanid pupae at high densities (407rr?) in Siberian peatlands. Of the numerous raised peat hu=ocks searched during early July 1990, many contained no tabanid pupae. These hu=ocks almost invariably contained the nests of an unidentified microtine rodent. Microtines such as lemmings will voraciously consume adult tabanids (Miller 1951), and it is possible that rodents are major predators of tabanid pupae in peatlands where they occur. In 1991, two larvae ofH. arpadi were colleeted which subsequently pupated • in the laboratory. One larva was colleeted on 11"June, pupated on 14 June, and 34 an adull Q emerged on 28 June. Another larva was collected on 19 June, pupaled • the same day, and an adull Q emerged on 30 June. Larvae and pupae were mainlained in the laboratory al 20· C, a temperature which was similar to that recorded immediately below the surface of peat hummocks in mid to late June. The observed 11-14 day pupation period for H. arpadi at Schefferville is considerably less the the 24 day pupation reported for this species (as Tabanus gracilipalpus) at Churchill (Miller 1951). The reason for this discrepancy is unclear, although pupae at Churchill were maintained in the laboratory at temperatures which were lower and considerably more variable (i.e., avg. 13" C, range 5-250 C) than those at Schefferville. Duration of tabanid pupation is very likely temperature dependent, although apparently no study has examined this. Miller's (1951) observation of the H. arpadi pupal period is based on a single • pupa, and my observation is based on only two pupae. Considerably larger samples of pupae are required to make meaningful comparisons. A singlelarva ofA. splzagnicola was collected on 18 June 1991 and pupated on 19 June; unfortunately it died shortly thereafter. AdultA. splzagnicola were collected 21 July - 4 August 1991, suggesting that (a) the pupation period for this. species is a month or more, (b) that the larva pupated prematurely, or (c) the pupation period of this species, like that of other tabanids, is approximately two weeks, and that adults were active for severa! days or weeks before they fust appeared in trap catches. 1 consiàer (c) most likely, since this species was only • rarely collected as adults at Iron Arm fen. 35 3. Eggs. • Although severa! searches were made during July of 1990 and 1991, no egg masses of any tabanid species were found on vegetation in Iron Arm fen. Reasons for this are unknown, although Miller (1951) reponed similar difficulty in locating tabanid egg masses in tundra meadows and ponds near Churchill. 4. Habitat Preferences. Numbers of larvae of each tabanid species colleeted at each site are presented in Table 4. 4.1. CIuysops spp. Nearly aIl Chrysops larvae (96%) were colleeted from sites where the water level was at or below the level of the substrate; very few C1uysops larvae of any • species were collected from sites where the substrate was submerged. Within this generalization, however, definite interspecific differences were apparent in preferred larval habitats. Larvae of C ater were colleeted ooly at Site 8, which was unique among the sites sampled in that its substrate consisted primarily of wet organic muck without overlying vegetation. While C ater larvae have not been previously reponed frorr. peaùands, they are known to favour substrates high in mud and organic matter, with litùe or no vegetation (Teskey 1969, Pechuman 1981, Burger et aL 1981). Site 8 was also preferred by larvae of C furcatus and C frigidus, although • severallarvae of these species were also colleeted at sites 1;1.,7 and 1,2,4 36 respectively. The preference of C furcatus and C frigidus larvae for a site with an • organic muck substrate is somewhat surprising. These species have previously been collected most commonly at sites with mossy substrates in open peatlands (Teskey 1969, Pechuman 1981, Burger et al.1981), although specimens of C furcatus have been collected from saturated clay soil (Teskey 1969), and C frigidus larvae are known to occur in sandy silt soil (Burger et al. 1981). Banbeau and Maire (1983a) collected larval C frigidus in both open and forested habitats in a bog near Trois-Rivières, Québec. Unlike larvae of the preceeding three C/uysops spp., larval C nigripes and C zinzalus were seldom colleeted at Site 8. Highest numbers of these species , were colleeted at Site 7, which was located in a jlark, or raised peat ridge through which a flow of water occurred. Relatively high numbers of larvae were also • collected at all other sites sampled, except for Sites 3 and 4, the two wettest sites. Teskey (1969) colleeted larvae of C nigripes from saturated moss on the banks of pools in tundra meadows. 4.2. Hybomitra, Atylotus spp. Of the 14 H. arpadi colleeted, 57% were colleeted from Site 3, the wettest site sampled, and a further 21% were collected from the other "very wet" site sampled. Unlike Chrysops larvae, larval H. arpadi appear to prefer submerged substrates of loose organic matter, although three mature larvae of this species were colleeted at Sites 1 and 2, where the substrate was relatively dry. While it is • possible these three larvae were in the process ofseeking pupation sites, it is also 37 possible that H. arpadi larvae are widespread within the peatla."ld. but are usually • too dispersed to be collected commonly. Even 50. these larvae apparently reach their highest densities (outside pupation sites) in aquatic sites. Teskey (1969) collected larval H. arpadi from saturated moss bordering shallow. water·filled depressions. It is impossible to say whether a preference for very wet substrata is shared by larvae of other horse fly species, as too few H. epistales, H. itasca, H. pechumani, H. zonalis and A. sphagnicola were colleeted during the present study for any conclusions to be drawn about their habitat preferences. In southem Québec, larvae of H. pec11umani were colleeted from forested and open habitats in an ombrotrophic bog, whereas H. epistales larvae were only colleeted from moss • bordering a pond in a minerotrophic fen (Baribeau and Maire 1983a). 4.3. Emergence Traps. No adult tabanids were colleeted in emergence traps during the summer in 1991, despite the faet that pupal exuviae had been colleeted in 1990 from the moss hummocks over which the emergence traps were placed. This is not surprising, in light of the low success rates that previons researchers have had in colleeting adult tabanids in emergence traps in northem Canada. Miller (1951) set ten émergence cages (2 m2 sampling area each) in peatlands near Churchill. From these, he colleeted only 20 adults of H. frontalis, H. afjinis, C nigripes; C frigidus, and C furcatus. He attributed bis low collection success, relative to . • known larval abundances, to the escape of adults thraugh holes in the traps, and 38 predation of adults within the traps by spiders and lemmings (Dicrostony;c • groenlandicus (Merrium». Maire (1984) reported a similar lack of success with emergence traps at Richmond Gulf in northern Quebee. Maire set out 40 emergence traps, each of which sampled 1 m2 of substrate, over moss hummocks in a bog. He collected only 12 adult tabanids, including specimens ofH. hearlei, H. lurida, C frigidus and an unidentified Atylotus species. He attributed his low sampling success to the low densities at which larvae were present in the peatland substrate, and concluded that emergence traps are no substitute for larval sampling. S. Seasonal Patterns of Larval Growth. In addition to larvae obtained as part of the study of the larval fauna and • distribution within the peatland, 70 tabanid larvae (Table 4, "OTHER") were obtained from sporadic collections from locations in the fen other than the eight study sites (e.g., moss hummocks during pupa searches, other invertebrate studies). These included 37larvae of Chrysops spp. (C ater, C frigidus, C furcatus, C nigripes, C zinzalus), 28 Hybomitra larvae (H. arpadi, H. epistates, H. itasca, H. peclzumani) and 5 A. sphagnicola. These larvae were measured, and their measurements included in the data set used to determine patterns of tabanid larval growth in the fen. • 39 5.1. Chrysops nigripes and C zinza1us. • One hundred and fourteen larval C nigripes and 150 C zinzalus \Vere colleeted and measured during the sununers of 1990 and 1991. Throughout both summers, larvae of each of these species consistently fell into three discrete length categories: 3.6-8.5 mm, 8.6-13.5 mm, and > 13.5 mm. This finding indicates that at least three distinct age-classes of larvae of these species occurring simultaneously within the fen. Mean lengths of larvae in each of the three length categories were plotted over the course of the two consecutive sununers. If the assumption is valid that each length category represents a larval cohort, then the resultant patterns suggest the limited larval growth in each cohon over the course of each summer. Seasonal cohon growth patterns are similar for both C nigripes and C zinzalus • (Figs. 6,7). Much of the fluctuation in the plotted lines is attributable to small sample size; on many sampling dates, particularly in 1990, one or no larvae were colleeted. Values for which no error bars are given were based u1?on a single observation. When plots of larval growth patterns are combined with.information concerning the known timing of adult flight seasons and presumed patterns of early!arval growth, pupation and egg incubation of each species, a generalized pieture of the life cycle of C nigripes and C zinzalus emerges suggesting that these species require 3-4 years to complete their development (Figs.8,9). Estimates of • the number of years required for larval development for T. nigrovittatus in salt 40 marshes (Meany et al 1976), and the tipulid Pedicia hannai Alexander in arctic • pools (MacLean 1973) have been obtained using similar combinations of data. First Summer: Adult female C nigripes and C zinzalus are apparently autogenous for the first gonotrophic cycle, so eggs of these species should be laid from a period extending from slightly before the beginning of the observed adult flight season until the end of the flight season. Oviposition in both species is therefore most concentrated during the last 2 weeks of July. Tabanid egg masses generally hatch 5-7 days after oviposition, and the first instar lasts only a few seconds (philip 1931, Lane 1974). Most second instar larvae should therefore have appeared in the fen during the last week of July and first week of August. Second instar tabanid larvae are approximately 1-2 = in length (philip 1931); no • larvae of this size were collected at all during the present study. Very small tabanid larvae are delicate creatures, relative to the larger, later-instar larvae, and second instar larvae may simply.have been too weak and susceptible to desiccation to be able to escape from drying peat samples in the dry extractor. Possibly, very smalliarvae II::lY have been overlooked, however great care was taken to avoid this. Chrysops larvae 3-4 = in length were collected on 18 August 1991, however these were thought be too large to have come from eggs laid in 1991. Despite the fact that no very young larvae were collected, limited growth must occur in a larva's first summer. Larvae 3.5-4.0 = in length (among the • smalIest collected) were taken in the spring, weil before the onset of adult flight 41 activity; these larvae must have been in at least their second summer. At the end • of the first summer, larval growth must halt in mid September to early October, when the fen substrate freezes solid. Larval development is arrested throughout the winter, until the substrate thaws approximately midway through the following May. Second and Third Summers: When the substrate thaws, larvae presumably again begin growing, and continue to do so throughout the summer months. During its second summer the larva grows slowly, increasing in size by 2-6 mm at MOst. Larval growth is again halted for the winter by the freezing of the larval habitat in mid September; larvae are approximately 8-10 cm in size at the end of their second summer. The following May the substrate again thaws, and larval growth and feeding can continue. As in the second SUŒ.IIler, larvae grow slowly, • increasing in size by only about 2-4 mm. By the time the substrate freezes at the end of their third summer, larvae are about 10-14 mm in length. Fourth and Firth Summers: Depending upon a larva's size when the substrate thaws at the beginning of its fourth summer, its development MaY follow one of two paths. Ifthe larva is sufficiently large, it grows for a brief periml, passes through a prepupal stage, pupates, and ultimately emerges an adult. Larvae of C nigripes can pupate when they reach 14-16 mm in length (Teskey 1969). Based upon the timing of adult emergence, and assuming that the duration of pupation is approximately two weeks, .larvae enter pupation over a period of approximately 3-5 weeks between late June and late July; the duration of pupation likely varies • somewhat from year to year and from species to species. Ifthe pattern of adult 42 abundance is taken as indicative of the pattern of larvae which entered pupation • two weeks previously, the majority of C nigripes and C zinzaIus larvae pupate in early to mid-July (Chapter 2, Figs. 5,6). Owing to the relatively short Chrysops flight season in the Schefferville area, it is assumed that a larva is physiologically programmed to only pupate within a prescribed 2-5 week period of the summer, in order that its emergence is co-incident with that of other members of its species. The eue to the larva may be photoperiodic, since the timing of the pupation period is relatively consistant between years. Regardless, if larvae during their fourth summer do not reach 14- 16 mm during the :'1-5 week period during which pupation is possible, then they do not pupate during their fourth summer. Instead, they continue growing until they reach their individual maximum possible sizes, up to approximately 18 mm in • length. The larvae continue to be active until the freeze-up in the fall. During the following (fifth) summer, further growth occurs ifthe larvae have not reached their maximum size, but larvae presumably pupate as soon as is photoperiodically possible. ...-- The collection of a C zinzaIus pupa on 5 July 1990, and the emergence of a female;from this pupa on 8 July, lend support to the presumed periods of -:. pupation and emergence presented in Fig. 9. 5.2. Other Chrysops spp. It is likely that a similar 3-4 year interval is also required for the • completion of the life cycles of C ater (Fig. 10), C frigidus (Fig. 11), and C 43 jurcatus (Fig. 12) in the Schefferville area, although relatively few larvae of these • tabanids were colleeted. When larvallength is plotted against collection date for these species, a pattern emerges which suggests severa! simultaneously-occurring larval age-classes, and to this extent resemble the patterns obtained for larval C nigripes and C zinzalus, 5.3. Summary: Chrysops Lire Cycles. To summarize, the life cycles of C nigripes and C zinzalus in the Schefferville area appear to require 3 to 4 years from egg to adult, dependirig upon the growth rate of individuallarvae; this may also be the case for other Chrysops spp. in the Schefferville area. At more southerly latitudes, Chrysops spp. require only one year (Terterian 1983, Logothetis and Schwardt 1948) or less • (Jones 1953) to complete their life cycles. The much longer life cycle of Chrysops spp. at Schefferville is a consequence of the very short period of the year during which the larval development cao occur. In Schefferville, the upper 5-10 cm of peatlands is only unfrozen for approximately 5 months, from mid May until early Oetober (Doug Barr, pers. comm.), and for only about three months of this period is substrate temperature greater than than Il)" C (Fig. 3). In New York State, where Logothetis and Schwardt (1948) found that C vittatus Wiedemann had a one year life cycle, the ground is usually unfrozen for 8 months from early April untillate November, and temperatures during this period average considerably • higher than Il)" C. In Florida, where C vittatus has 2 generations/year (Jones '. 44 1973), the ground does not freeze at aIl, and larval development continues year- • round. Life cycles in which multiple years are spent as larvae are typical of a number of inseet groups which live in habitats where the growth season is very short, and short periods of growth are separated by long periods of dormancy (Danks 1992). Chironomus spp. (Diptera: Chironomidae) in aretic pools may spend up to seven years in the larval stage (Butler 1982); moreover, iflarvae have not reached a size suitable for pupation by the beginning of a given summer, they will remain as larvae until the following summer, thereby adding an additional year to their life cycles (Danks and Oliver 1972). A similar pattern ofvariable . life cycle length, termed "cohort-splitting", was been observed in TIpula sacra Alexander (Diptera: Tipulidae) in temperate beaver ponds (Pritchard 1980), and • also occurs in aretic Lepidoptera (Downes 1965) and subaretic ground beetles (Coleoptera: Carabidae)(Kaufman 1971). Life cycles ofvariable length have also been reported previously for tabanid species which typically have one-year life cycles; an additional year is added to the life cycles of larvae which are not large enough to pupate durin,g their second summer (Tashiro and Schwardt 1953, Magnarelli and Anderson 1978). 5.4. Annual Variability ofChrysops Larval Cohorts. Because of the prolonged life cycle, larvae from severa! generations • obviously co-exist in the soil at any given time. During each of the 2 summers of 45 study, C nigripes and C zinzalus larvae from five different cohorts were • presumably present in the fen substrate: (1) 1-3 mm first-summer larvae, present only during and after the adults emerge in mid-July (no individuals belonging ta this cohort were colleeted); (II) <85 mm second-summer larvae, present throughout summer; . (III) 85-135 mm third-summer larvae, present throughout summer; (IV) >135 mm, fourth-summer larvae which will either pupate early in the summer, or attain pupation size tao late to emerge that year; (V) >135 mm, fifth-summer larvae awaiting pupation, present only in early summer, most pupate by late July (Cohorts IV and V could not be differentiated). Unlike what might be predieted from a survivorship curve, the relative • sizes of successive cohorts of C nigripes differed considerably between 1990 and 1991 (Fig. 14). In 1990, the majority of larvae colleeted were second-summer, whereas in 1991, most larvae were in their third summer. The predominance of the 1990 third-summer cohort was refleeted in the large size of the fourth-summer cohort in 1991. A similar pattern was found with C zinzaIus larvae, except that in this case the third-summer cohort was the largest in 1990, whereas the fourthjfifth-summer cohort was largest in 1991 (Fig. 15). This high variability among cohort sizes, and the refleetion of one year's larval cohort structure in the cohort structure the next year, suggests that larval tabanid populations in a given peatland may vary cyclica1ly from year ta year. A • year during which large larvae are plentiful should be followed by a year in which 46 adults are plentiful. Many adults potentially lay many eggs, beginning the cycle • anew. Altematively, a year in which few mature larvae are present should be followed by a year in which few adults emerge, and thus few eggs are laid. The abundance or paucity of adults in a given year could potentially be reflected in the abundance of adults three years hence. This interpretation is somewhat oversimplified, however. There is some evidence that larval populations in each cohon vary among peatlands in a given area (see Chapter 2), so that even if adult emergence from a given peatland is low in a given year, many females may be present because of a "good" emergence from a nearby peatland. If a summer's weather is dominated by conditions unfavorablef~r!!i~ht activity by adult tabanids (e.g., host-seeking, mating; see Chapter 3), oviposition in a peatland may be low despite large numbers of flies • emerging there. Finally, because of "cohon-splitting", the presence of a large number of large larvae in the fen substrate does not necessarily mean that they will become adults the next year. Rather, they may or may not pupate, depending upon their individual rates of development Since development rate is dependent . upon substrate temperature, a cold spring in a given year, or a cold summer in a preceeding year, could lead to a reduction in the number of larvae pupating, whereas a warm spring could result in the pupation of more larvae. 5.5. Hybomitra spp. Since only 25 H. arpadi larvae were collectecl, it was difficult to determine • the annual growth pattern of this species (Fig. 13). Assuming that Chrysops spp. 47 in the Schefferville area require from 3 to 4 years to complete development, • however, and that Hybomitra spp. larvae pupate at considerably greater sizes than those of Chfysops spp. (e.g., 20-30 mm vs. 12-18 mm)(Teskey 1969), it seems reasonable to assume that Hybomitra spp. require at least as long as Chrysops spp. to complete their larval development, and perhaps longer. 5.6. Growth of Captive Chrysops and Hybomitra Larvae. Throughout the summer of 1990, 109 tabanid larvae were maintained in captivity for up to 60 days. Even though larvae were fed and measured weekly, in no case was larval growth apparent in the terms of changes in head capsule or body lengths. While it was expectèd that larvae collected when fully grown would show no further growth in captivity, it was somewhat surprising that smaller larvae • did not increase in size. Teskey (1969) observed growth in captive tabanid larvae which were fed severa! blowfly maggots per week, but he did not deterrnine how many of the maggots the tabanid larvae aetually devoured. Burger (1977) found that growing Hybomitra and Tabanus larvae cODSUfiied 3-10. Possibly, the captive tabanid larvae larvae during the present study were not fed often enough to provide for any more than their maintainance requirements. • 48 • LITERATURE CITED: Andreeva, R.V. 1989. The morphological adaptations of horse fly larvae (Diptera: Tabanidae) to developmental sites in the Palearctic Region and their relationship to the evolution and distribution of the family. Cano J. Zool. 67: 2286-2293. Bailey, N.S. 1948. A mass collection and population survey technique for larvae of Tabanidae (Diptera). Bull. Brooklyn EntamaI. Soc. 43:22-29. Baribeau, L. and A. Maire. 1983a. Spatial distribution of Tabanidae (adults and larvae) in two bogs of southern Québec. Mosquito News 43: 24-29. Baribeau, L. and A. Maire. 1983b. Abundance and seasonal distribution of Tabanidae in a temperate and a subarctic locality of Québec. Mosquito News 43: 135-143. Booth, T.W., Jr. 1968. The availability and utilization of the foods of the common snipe (Capella gallinago de7icata) in the rice-growing region of southwestern Louisiana. Unpublished M.Sc. Thesis, Louisiana State University, Baton Rouge. Burger, J.F. 1977. The biosystematics of immature Arizona Tabanidae (Diptera). Trans. Amer. Entom07. Soc. 103: 145-258. Burger, J.F., D.J. Lake, and M.L. McKay. 1981. Larval habitats and rearing of sorne common Chrysops species (Diptera: Tabanidae) in New Hampshire. • Proc. EntamaI. Soc. Wash. 83: 373-389. Butler, M.G. 1982. A 7-year life cycle for two Chironomus species in arctic Alaskan tundra ponds (Diptera: Chironomidae). Cano J. Zo07. 60: 58-70. Chvala, M., L. Lyneborg, and J. Moucha. 1972. The horse flies of Europe. Ent. Soc. Copenhagen Publ., pp. 38-44. Oanks, H.V. 1992. Long life cycles in insects. Cano Entom07. 124: 167-187. Oanks, H.V., and O.R. Oliver 1972. Seasonal emergence of sorne high arctic Chironomidae (Diptera). Cano Entom07. 104: 661-686. Oownes, J.A. 1965. Adaptations of insects in the arctic. Ann. Rev. Entom07. 10: 257-274. Dukes, J.C. and K.L. Hays. 1971. Seasonal distribution and parasitism of eggs of Tabanidae in Alabama. J. Econ. Entom07. 64: 886-889. Dukes, J.C., T.D. Edwards, and R.C. Axtell. 1974. Distribution of larval Tabanidae (Diptera) in a Spartina a7ternif70ra salt marsh. J. Med. • Entom07. Il: 79-83. 49 Ellis, H.C. and K.L. Hays. 1973. Population densities of tabanid larvae in two farm pond habitats in east central Oklahoma. Florida Entomol. 56: • 11-14. Fairchild, W.L., M.C. O'Neill, and D.A. Rosenberg. 1987. Quantitative evaluation of the behavioral extraction of aquatic invertebrates from samples of sphagnum moss. J. N. Am. Benthol. Soc. 6: 281-287. Freeman, J.V., and E.J. Hansens. 1972. Collecting larvae of the salt marsh greenhead Tabanus nigrovittatus and related species in New Jersey: comparison of methods. Environ. Entomol. 1:653-658. Gingrich, R.E. and R.A. Hoffman. 1967. Abundance and survival of tabanid larvae in effluent from a dairy barn. Ann. Entomol. Soc. Amer. 60: 72-74. Graham, N.L. and J.G. Stoffolano Jr. 1983. Relationship between female size, type of egg mass deposited, and description of the oviposition behavior of the sibling species Tabanus nigrovittatus and T. simulans (Diptera: Tabanidae). Ann. Entomo7. Soc. Amer. 76: 699-702. Hill, G.F. 1921. The bionomics of Tabanus aprepes and other Australian Tabanidae. Bu71. Entomo7. Res. 12: 41-61. Hine, J.S. 1906. Habits and life histories of some flies of the family Tabanidae. U.S.D.A. Bur. Entomol. Tech. Ser. 12:19-37. Isaac, P.V. 1924. Papers on Indian Tabanidae. 1-111. M~m. Dept. Agr. India, • Entomo7. Ser. 8: 53-62. Isaac, P.V. 1925. Papers on Indian Tabanidae. VI-VII. Hem. Dept. Agr. India, Entomol. Ser. 8: 93-109. Jackman, R., S. Nowicki, D.J. Aneshansley, and T. Eisner. 1983. Predatory capture of toads by fly larvae. Science 22: 515-516. Jackson, J.O. and B.H. Wilson. 1965. Observations on some predators of horse fly and deer fly (Diptùra: Tabanidae) eggs in Louisiana. Ann. Entomo7. Soc. Amer. 58: 934-935. James, H.G. 1953 .. Natural control of Tabanidae (Diptera) in the region of Churchill, Manitoba. Ann. Report Entomo7. Soc. Ontario 82: 70-74. James, H.G. 1963. Larval habitats, development, and parasites of some Tabanidae (Diptera) in southern Ontario. Cano Entomo7. 95: 1223-1232. Jones, C.M. 1953. Biology of Tabanidae in Florida. J. Econ. Entomo7. 46: 1108-1109. Kaufmann, T. 1971. Hibernation in the arctic beetle, Pterostichus • . brevicornis, in Alaska. J. Kansas Entomo7. Soc. 44: 81-92 • 50 Khan, N.H. 1951. The bionomics of tabanid larvae (Diptera). J. Bombay Nat. • Hist. Soc. 51: 21-28. King, H.H. 1910a. Some observations on the bionomics of Tabanus par Walker and Tabanus taeni07a Pal. de Beauv. Bu77. Entom07. Res. 1: 99-105. King, H.H. 1910b. Some observations on the bionomics of Tabanus ditaeniatus Macquart and Tabanus kingi Austen. Bu7l. Entom07. Res. 1: 265-274. Knudsen, A.B. and D.M. Rees. 1968. Methods used in Utah for sampling tabanid populations. Mosquito News 28: 356-361. . Lane, R.S. 1976. Density and diversity of immature Tabanidae (Diptera) in relation to habitat type in Mendicino County, California. J. Med. Entomol. 63: 283-291. Leprince, D.J. and M. Bigras-Poulin. 1988. Seasonal variation in body size and the relationships amongst body size, parity, and fecundity of host-seeking Tabanus quinquevittatus females (Diptera: Tabanidae). J. Med. Entomol. 25: 105-110 Leprince, D.J. and P. Jolicoeur. 1986. Annual, intra-, and interspecific variations in body size and potential fecundity of some Tabanus species (Diptera: Tabanidae). Cano Entom07. 118: 1265-1272. Lewis, L.F. and C.M. Jones. 1955. The biology of tabanids in the Mississippi Yazoo Delta. J. Econ. Entomol. 48: 609-610. • Logothetis, C. and H.H. SchwarJt. 1948. Biological studies on the horse flies of New York. J. Econ. Entomol. 41: 335-336. Lutta, A.S. 1970. Horse flies (Diptera: Tabanidae) of Karelia. 303 p. Leningrad. [In Russian). Lutta, A.S., and K.I. Bykova. 1982. Horse flies of the european north of the U.S.S.R. Leningrad. 184 p. [In Russian). MacLean, S.F. Jr. 1973. Life cycle and growth energetics of the arctic crane fly Pedicia hannai antenatta. Oikos 24:436-443. Magnarelli, L.A., and J.F. Anderson. 1978. Distribution and development of immature salt marsh Tabanidae (Diptera). J. Med. Entom07. 14: 573-578. Magnarelli, L.A., and J.F. Anderson. 1979. Oogenesis and oviposition in Chrysops at7anticus (Diptera: Tabanidae). Ann. Entom07. Soc. Amer. 72: 350-352. Magnarelli, L.A., and J.F. Anderson. 1980. Parasitism of Tabanus nigrovittatus immatures (Diptera: Tabanidae) by Trichopria sp. • (Hymenoptera: Scelionidae). J. Med. Entom07. 17: 481-482. 51 Magnarelli, L.A., D.J. Leprince, J.F. Burger, and J.F. Butler. 1982. Oviposition behavior and fecundity of Chrysops cincticornis (Diptera: • Tabanidae). J. Med. Entom07. 597-560. Martin, J.E. 1977. Collecting, preparing, and preserving insects, mites, and spiders. The insects and arachnids of Canada. Part 1. Ministry of Supply and Services, Canada. Ottawa, Ontario. 182 p. McElligott, P.E., and T.D. Galloway. 1991. Seasonal distribution and parity of host-seeking horse flies (Diptera: Tabanidae) from a pasture near Seven Sisters, Manitoba. Cano Entom07. 123: 361-370. Meany, R.A., 1. Valeila, and J.M. Teal. 1976. Growth, abundance, and distribution of larval tabanids in experimentally fertilized plots on a Massachusetts salt marsh. J. App7ied Ec07. 13: 323-333. Miller, L.A. 1951. Observations on the bionomics of sorne northern species of Tabanidae (Diptera). Cano J. Zo07. 29: 240-263. National Wetlands Working Group. 1988. Wetlands of Canada. Sustainable Development Branch, Environment Canada. Ottawa, Ontario. 452 p. Olsufjev, N.G. 1949. On the horse fliescof the northern part of the Baraba forest-steppe and sorne methods of their control. Voprosy kraevoi, obshcheii eksperimenta7'noi parazit070gii (Moskva) 6: 34-58. [In Russian]. Olsufjev, N.G. 1977. Horse flies, Family Tabanidae. Dipterous Insects,Vol. • VII, No. 2. Fauna SSSR, novaia seriia 113: 1-434. [In Russian]. Orminati, S.A., and E.J. Hansens. 1974. The biology of Tabanus 7ine07a 7ine07a F. Ann. Entom07. Soc. Amer. 67: 937-939. Otsuru, M. and S. Ogawa. 1959. Observations on the bite of the tabanid larva . c in paddy-fields (Diptera: Tabaniâae). Acta Medica et Bi070gica 7: 3F5û. Pechuman, L.L. 19B1. The horse flies and deer flies of New York (Diptera: Tabanidae). Second edition. Search Agricu7ture 18: 1-71. Pechuman, L.L., H.J. Teskey, and D.M. Davies. 1961. The Tabanidae (Diptera) of Ontario. Proc. Entom07. Soc. Ontario 91: 77-121 • .Pechuman, L.L., D.W. Webb, and H.J. Teskey. 1983. The Diptera or true flies of Illinois. 1. Ta~anidae. 177. Nat. Hist. Surv. Bu77. 33(1): 1-121. Philip, C.B. 1931. The Tabanidae (Diptera) of Minnesota, with special reference to their biology and taxonomy. Univ. Minn. Agric. Exp. Sta. Tech. Bu77. 80: 1-132. Pritchard, G. 1980. Life budgets for a population of Tipu7a sacra (Diptera: Tipulidae). Ec07. Entom07. 5: 165-173 • • .~ - , 52 Roberts, R.H. 1966. Biological studies on Tabanidae. 1. Induced oviposition. • Mosq. News. 26: 435-43B. Roberts, R.H. 1980. The effect of temperature on the duration of oogenesis and embryonic development in Tabanidae (Diptera). J. Med. Entom01. 17: 8-14. Roberts, R.H., and R.J. Dicke. 1964. The biology and taxonomy of sorne immature nearctic Tabanidae (Diptera). Ann. Entom01. Soc. Amer. 57: 31-40. Roth, A.R., and A.W. Lindquist. 1948. Ecological notes on the deer fly at Summer Lake, Oregon. J. Econ. Entom01. 41: 473-476. Schomberg, O. 1952. Larval habitat of Tabanus su1cifrons in Oklahoma. J. Econ. Entom01. 45: 747. Schomberg, O., and D.E. Howell. 1955. Biological notes on Tabanus abactor Phil. and T. equa1is Hine. J. Econ. Entom01. 48: 618-619. Segal, B. 1936. Synopsis of the Tabanidae of New York, their biology and taxonomy. J. N.Y. Entom01. Soc. 44: 51-76, 125-154. Sperry, C.C. 1940. Food habits of a group of shore birds; woodcock, snipe, knot, and dowitcher. U.S. Dept. Interior, Wildl. Res. Bull. No. 1., 55 pp. Stone, A. 1930. The bionomics of sorne Tabanidae (Diptera). Ann. Entom01. • Soc. Amer. 23: 261-304. Tashiro, H. and H.H. Schwardt. 1949. The biology of the major species of horse flies in central New York. J. Econ. Entom07. 42: 269-272. Tashiro, H. and H.H. Schwardt. 1953. Biological studies of horse flies in New York. J. Econ. Entom07. 46: 813-822. Terterian, A.E. 1985. Bioecological peculiarities of immature stages of horse flies in the U.S.S.R. (Diptera: Tabanidae). Myia 3: 485-514. Teskey, H.J. 1962. A method and apparatus for collecting larvae of Tabanidae (Diptera) and other invertebrate inhabitants of wetlands. Proc. Entom01. Soc. Ontario 92: 204-206. Teskey, H.J. 1969. Larvae and pupae of sorne Eastern North American Tabanidae (Diptera). Mem. Entom01. Soc. Cano 63: 1-147. Teskey, H.J. 1983. A revision of the Eastern North American species of Aty10tus (Diptera: Tabanidae) with keys to the adult and immature stages. Proc. Entom01. Soc. Ontario 114: 21-43. Teskey, H.J. 1990. The Horse flies and deer flies of Canada and Alaska (Diptera: Tabanidae). The insects and arachnids of Canada. Part 16. • Ministry of Supply and Services; Canada. Ottawa, Ontario. 381 p. 53 Thomas, A.W. 1971. The biology and taxonomy of immature Tabanidae from Mendocino County, California, with an autecological study of Chrysops • hirsutica77us Philip (Diptera: Tabanidae). Unpublished Ph.D. Dissertation. University of California. 266 p. Thompson, P.H. 1970. Larval Tabanidae (Diptera) of the Great Swamp, New Jersey. Ann. Entom07.Soc. Amer. 63: 343-344. Thompson, P.H. 1971. Larval Tabanidae (Diptera) of the Patuxent Wildlife Research Center, La~rel, Maryland. Ann. Entom07. Soc. Amer. 64: 956-957. Thompson, P.H., J.W. Holmes Jr., P.C. Krauter, C.M. Raney, and M.J. Clay. 1979. Rearing of Texas Tabanidae (Diptera) II. Mass production of Tabanus nigrovittatus Macquart eggs and larvae. Southwest. Entom07. 4: 224-230. Thompson, P.H., B.F. Hogan, and H. Del Var Petersen. 1980. Rearing of Texas Tabanidae (Diptera). III. Trapping, survivorship, and limited rearing of Hybomitra 7asiophtha7ma (Macquart). Southwest. Entom07. 5: 191-195. Tidwell, M.A., and K.L. Hays. 1971. Oviposition pref,"~ences of sorne Tabanidae (Diptera). Ann. Entomo7. Soc. Amer. 64: 547-549. Tuck, L.M. 1972. The snipes. Canadian Wildlife Service Monograph Series, No. 5. 429 p. Twinn, C.R., B. Hocking, W.C. McDuffie, and H.F. Cross. 1948. A preliminary • account of the biting flies at Churchill, Manitoba. Cano J. Res. (Sec. D): 26: 334-357. Wall, W. and H. Jamnback. 1957. Sampling methods used in estimating larval populations of salt marsh tabanids. J. Econ. Entomo7. 50: 389-391. Wilson, B.H. 1969. Tabanid larval habitats and population densities in an alluvial area of southern Louisiana. Ann. Entomo7. Soc. Amer. 62: 1203- 1204. . Zapekina-Dul'keit, 1.1. 1966. On numbers, breeding sites, and development of horse fly larvae in the "Stolby" Preserve. pp. 55-56. ln Voprosy zoo70gii. Materia7y k III soveshchaniiu zoo70gov Sibiri. Tomsk. [In Russian). Zapekina-Dul'keit, 1.1. 1969. Horse flies and other blood-sucking dipterans of the "Stolby" Preserve. Voprosy entom07ogij. Trudy Gosudarstvennogo zapovednika "St07by" 7: 4-105. Krasnoiarsk. [In Russian). • 54 Table 3. Average percent cover of major vegetation and exposed substrate types, relative wetness, pH, and years sampled at eight sampling sites in Iron • Arm fen, 1990 and 1991. Site Vegetation su: Substrate Type 1 2 3 4 5 6 7 8 Sctrgus çr~'ptto,us 92 37 68 47 13 Ertophorum augustifolt um 22 Friophor'um chamts,onts + 21 ~ spp. + 75 5 62 52 5 Andromeda glaucophyl1a 7 + Myrtea gale 23 22 25 5 Menyanthes trtfoltata + Smllac1nll trtfolia + Hosses (e,g, Sphognum spp.) 7 10 48 2 Loose flocculent • organ fC rMtter 40 lIet mud 67 Open \lISter 95 7 7 15 Relative Wetness Category 0 0 VII VII 0 0 Il Il Average pH (. stdorr) 5.6 .0,1 5,4 .0,0 5.5 .0.0 5.7 .0,1 5.9 .0.0 5.4 .0.1 5,6 .0.1 5.S.0.0 Years Sompled: 1990 + + + + 1991 + • + + + + + + "." - few plants of th's spec;es present Relative IIetness categories: 0 (dry) - lIater table 3-5 cm below surface of substrate Il (wet) - lIater table at surface of substrate • VII (very wet) - lIater table above substrate surface. vegetation present • D • • 55 Table 4. sam~lin~ effort and total number of larvae collected at eight sampling sites in Iron Arm fen, 1990 (Si es -4) and 1991 (Sites 1-8). SITE 1 2 3 4 5 6 7 8 TOTAL OTHER2 or{ Ext. Samples' 22/25 22/25 22/25 22/25 24 25 25 25 287 We Ext. Samples 42/24 42/24 42/24 42/24 24 24 24 24 360 Chrysop.5 ater 17 17 4 C. frigidus 0/1 1/1 2 ID 15 1 C. furcatus OZ! 0/2 0/1 14 17 9 C. ntgl-ip'es 10/17 14/11 213 6 16 28 1 108 6 C. zinzalus 12/21 8115 . 4/2 12 30 34 7 145 5 Unknown Chl'Ysops 1/9 0/12 1/0 2/0 3 4 16 6 54 12 Total Chrysopsinae 23/49 23/41 1/0 8/6 21 50 80 55 357 37 Hybomitra arpadi 1/0 1/1 5/3 2/1 14 11 H. epfstates 1 1 2 1 H. itasca 0/1 1 5 H. pechumani 2/0 1 1 1 5 1 H. zonalis 1/0 1 2 Unknown Hybomitra 1/3 0/1 1/5 1/4 3 2 2 23 10 Atylotus sphagnfco7a 1/0 1 5 Total Tabaninae 5/3 1/2 7/9 3/5 6 4 3 2 48 33 Total Tabanidae 28/52 24/43 8/9 11/11 27 54 83 57 406 70 , 1990/1991 numbers of samples taken and numbers of larvae collected for sites 1-4; 1991 ooly for Sites 5-8. 20THER = larvae collected from sites other than Sites 1-8, e.g., from moss hummocks • 56 Figure 1. Location of Schefferville, Quebec, and approximate location • of study area. • • • <~ Lo'~~ :. .. .4".07 ; !: ,,\ , '\:.:~ •. Nu",!:ers ;,.,Jiccte loeclio", o( sampt~ o 5 10 15 r~"~I~1 1 , , ~.'-:!.-:.: Lofees' • 1t11~,"e"" . . 16·}j S"'omp ondfI:Jn ..wooded mUlfceg I~HHmllonrJ0&0." 700 me'tes Comp,tec1 in 1978 (rom sfu:el. 23) end 230 oF rlu: Conoc/;on DC"porltnenr oF Mincs onrl TC"chnicol Sur",t:'7J (J963J • : • 57 Figure 2. Ma~ of Iron Arm fen, 20 km ne Schefferville Quebec, inâicating moi sture regimes and locations of sampling sites • in 1990 (Site 1-4) and 1991 (Sites 1-8). • • • rPOW 'li,.. r ••• La...... oP o .at... _ •• ...r t ...... 1.. C ....1.•• 1...... 1_ ••rtoco • _te. t •••••" _r'.co • .~"'Ill. _te. wlU...... tlnll • ..,.... _ ""'...... tlo• • • 58 Figure 3. Mean (± standard error) weekly temperature in top 10 cm of substrate in Iron Arm fen during summers of 1990 and 1991. Temperature was recorded at 4 sltes ovar Il weeks in 1990 and • at 8 sites over 12 weeks in 1990. • --, .'.";:; • 1 1 o en en en en - -::>.. ::> -<:'" Hjl"4'" 1 ...... 'loi, ,, • ,, f-6t ...... QI c: ::> ""') ..,o lfl o lfl o o C'l C'l - - • <.:: • 59 Figure 4. Schematic diagram of apparatus used in (a) wet extraction, • and (b) dry extraction. • • A WET EXTRACTION APPARATUS - 10aWJgltbUb ,.. plastic container _bolh • B DRY EXTRACnON APPARATUS llO W JgltbUb _hmo wnl1'lOlh 't\====-:J;-~- • • 60 Fi9ure 5. Patterns of pubescence on larvae of (a) Chrysops presumed zinzalus Philip, Cb) C. nigripes Osten Sacken variant collected at Schefferville, and (c) C. nigripes described by • Teskey (1969). • • A ,.' id' Chrysops zlnzalus Philip (presumed) B • Chrysops nlgrlpes Osten Sacken - Schefferville Variant .. , ~!: c ~ 1 '-1':~ ... ;:~ ..~ ~. • ,. iï ~i ~ ·[; .\ ! 1 \' • 1 \ or .; fi: ':i ..:.~ .!~ Chrysops nlgrlpes Osten Sacken - Typlcal Form • • 61 Fi9ure 6. Weekly mean (± standard error) lengths of larval Chrysops nigripes Osten Sac ken in three size categories collected from • all sites in Iron Arm fen in 1990 and 1991. • • • • .... • V! :> Cl :> I+l • < • • lei ~ ~ ~ ~ al :> al ., ~ lei lei ~ ~ • w I-e-l z • .,:> I+l • 1 • .... • V! I+l :> • Cl :> • lei < • • • • 0 ~ ~ al • al • .,:> • zw • :> • ., 0 Cl) N CO V N - 0 • (WW) 4l6ua1 IOlU01 uoaYi '. • Figure 7. Weekly mean(± standard errorl lengths of larval Chrysops ~resumed zinza7us Philip in three size categories collected • from all sites in Iron Arm fen in 1990 and 1991. • • tel • I+! I-+-l 1- Vl .-l :J Cl • :J • .....-l < .-l • ...-..... tel ~ ...-..... I+! -al ..,:J al I-+-l lei - I-+l I+! • ...... W • Z :J.., ...-t 1 • 1 • 1- Vl • :J Cl • • :J I-+l < • • a • ~ al • • ..,:J al • • • • • p a (l) C'l !XI -:- C'l - - a • (ww) 4l6ual IOlUol uoan • Fi9ure 8. Schematic representation of presumed life cycle of Chrysops nigripes Osten Sacken in Iron Arm fen. indicating likely • growth patterns of 1986-1991 larval cohorts. • , ~ 4 ~ , ~ 1 o 1 1 \ '" 1 0 , ~ , "u , 1 , , 1 ;;; , lt~ \ :: , 1 , l- l/l -r I 1 => 1-0-1 ...J.. 0 1 1 => 1 < • " 1 1 51 1 \ " 1 !:II \ "... 1 "1 ~,' 1 1 .. , 1 ..L.. 1 1 ~ Cl , 1 ...,=> Cl ...L- I - 1 \ \ 1 W 1 Z 1 ...,=> 1 , 1 ~ 1 1 1 l' 1 1 l ' \ 1 , 1 , 1 1 t , 1 1 1 1 1 , Zw 1 N li: ..a< li: 1 .. , 0 0 1 0 a x x 0' '" 1 Il: 0 0 1 SI "0 1 .... ., 1 "1 1 "... " 1 .. "0 a ...... 1 .. 1 Z .. .. 1 .., .. , , 1 a=> 1 Il: , 1 1 Cl , , • 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 l- 1 , l/l 1 I 1 => , Cl 1 => , 1 < 1, 1 -.- 1 1 1 , 1 .. 1' 0 , 1 s," ~ Cl => Cl !:l' 1 ..., ~I •1 il , ..L.. "1 1 , 1 1 ...L- I ....~ 1 w ..~~ , Z .. a< ~~ .,0 1 =>..., .. x 1 0 ~ " 0 co C'l IX) ...,. 0< C'l :/- - - • (ww) 4l6u31 )DI\JDl UD3n ? • Figure g. Schematic representation of presumed life cycle of Chrysops presumed zinza7us Philip in Iron Arm fen, indicating likely • growth patterns of 1986-1991 larval cohorts. • A ~ 1 , ft. ~~ 1 , , 0:1 1 ~ 1 , 81 1 - 1 \ ~ t- ~ , 1 , 1 1 1 1 t 1 , 1 Z 1 W ' ... 1 1 ... 1 N li: ~I li: 0 1 0 1 :31 0 X 1 x, X 1 Il:: o 0 S' L0- U 18. U 1 UI ~ 0 ...co 1 :_ co 1 gl Z .. 1 ~I .. 1 :1 ::J 1 1 1 0 1 Il:: , 1 1 , t.:l , 1 1 1 1 1 1 1 ~ • , \ 1 , , 1 , \, , 1 , , 1 \ , \ 1 1 \ \ 1 1 \ \ 1 1 \ , 1 1 , \ , , \ \ , 1 , l- I 1 VI ...,.... 1 ::J 1 + t.:l 1 1 ::J 1 1 , < 1 1 1 .. 1 ~I 1 ..." 1 !:l' 1 0 .. 1 ....L.. ~I ~ Cl ::> Cl 1 ..., 1 - 1 ....L.. \\ 1 , W ""',... , , , \ Z 0: , 1 \ ..co 0 ..,::> ~ '1 \ 8 ...- \ 0 ". • Figure 10. Weekly mean (± standard error) lengths of larval Chrysops ater Macquart in three size categories collected fro~ all • sites in Iron Arm fen in I990 anâ 1991 . • • 20 .... 0 0 16 C- o ,... 0 E ~ E '-' 0 ~ 12 c- • C'l -C CIl -1 • 0 • e ~ e 8 t- -1 eC CIl :::l: • f- 4 o I--...l-'----:'L.-_.l-'-I--'L.--.l-'-1'_+_..1-'_...l..'-.l-'-l June July August • 1990 o 1991 • • DD Figure Il. Weekly mean (± standard error) lengths of larval Chrysops frigidus Osten Sacken in three size categories collected from • all sites in Iron Arm fen in 1990 and 1991 . • • 20 ,.. T o 16,.. 2 î ...... c E ...... E o • .r:. ~ 12 - C' c: CIl -l o • ~ 8,.. -l c: o o CIl ::l: • 41- o 1-_.1.-'_...l-'---1'-+-1..-'-.1.-'_...l-'-j_...J'L..-_I..-'_.I.-'--1 June July August • 1990 o 1991 • • DI Figure 12. Weekly me an (± standard error) lengths of larval Chrysops furcatus Walker in three size categories collected from all • sites in Iron Arro fen in 1990 and 1991 . • • 20 - 0 0 16 ~ 0 ~ 0 ...... 0 0 E !:! 0 ...... ,E 0 • .s:: 12 C' - -c: 0 QI .J 0 0 0 0 ! ~ 0 8 1- .J 0 • • 0 c: 0 0 • CIl • ;:l; 4 f- o f-_..l-'---J'i.-_1....-'--+_"';'L-_l....'---J'_--1-_l....'_...L'_1....-'--1 June July August • 1990 o 1991 • • Figure 13. Len9ths of larval Hybomitra arpadi Szilady collected from all • sites in Iron Arm fen in 1990 and 1991 . • a ~ • en en en en ~ ~ • 0 0 ..... • CIl ::J • 0) • «::J • 0 •• • 0 • •• ~ ::J """') • •• 0 • 0 0 • 0 CD • C ::J 0 • """') 1 1 1 1 1 lt) o o C\l C\l (ww) Lll5u8l relüel • • 6') Figure 14. Percentage of Chrysops nigripes Osten Sacken larvae from Iron Arm fen ln 1990 and 1991 which fell into each of 3 size • categories. • • C\J Cl n Il • t: R 1: Q) al - ~ () Cl ... - II- a.Q) 2 0 Cl) C\J Il t: R t: 0 -Q) al () al ~ ... • - a.Q) -- 2 0 E E E E E Il) E Il) C'Ï Il) C'Ï - cD -1\ Il) V 0) cD .~ ~ c: f; ~ Cl t: ~ Q) t3 ...J • • 70 Figure 15. Percentage of Chrysops presumed zinzalus Philip larvae from Iron Arm fen in 1990 and 1991 fen which fell into each of 3 • size categories. • • v ~ C\l ~ 11 c: l'l ~ ë m Q) m !il Ü ~ ~ Q) -- C- S! 0 co C\l 11 c: l'l c: 0 -Q) m m !il Ü ~ ~ Q) C- • -- S! 0 E E E E E III E III ci III ci ~ ai ~ III V /1 ai ~'" .~ '"g. ~ c: ~ Q) ..c: ....J Ü • • ïl CHAPTER2 Diversity, Seasonal Activity, and Seasonal Changes in the • Gonotrophic Age Structure of Host-Seeking Tabanidae from Two Peatlands Near Schefferville, Québec. • • 1. LITERATURE REVIEW 1. DISTRIBUTION OF ADULT TABANIDS. The distribution of adult ,:tbanids, on both geographical and local scales, is dependent upon a wide variety of interacting factors, incIuding larval habitat type, choice of host-seeking site by females, choice of mating site by males, and dispersal of flies from emergence sites. 1.1. Larval Habitat vs. Adult Distribution. One of the most important factors goveming local abundance of adult tabanids is the presence or absence of suitable larval habitat. Some tabanid species have very specific habitat requirements. Hybomitra pediontis (McAlpine), • for example, is an alkaline-lake breeder, and its geographic distribution is cIosely tied with the distribution of these la.\:es (Hanec and Bracken 1964). Sinùlarly, Tabanus nigrovittatus Macquart is restricted to coastal salt marshes along the Atlantic seaboard (Joyce and Hansens 1968, Rockel and Hansens 1970, Dale and Axtell 1976). These species may be common in areas with suitable larval habitat, but rare elsewhere. Other tabanid species are more widespread, either because they have more catholic larval habitat requirements, or because their panicular larval habitat is geographically common. Hybomitra lasioplztlzalma (Macquart) is common throughout eastem North America (Pechuman et ai. 1983), and is known to breed • in a wide variety of habitats (Teskey 1969). Hybomitra frontalis (Walker) breeds ï3 exclusively in peatland habitats (Teskey 1969). but is widespread and abundam in • areas of boreal and subarctic North America where peatlands are common. 1.2. Habitat Selection by Host-Secking Females. Locally, the numbers of host-seeking female tabanids encoumered may vary considerably among different habitat types (Davies 1959, Hancc and Brackcn 1964, Roberts 1969, Smith et aL 1970, Golini and Wright 1978, Leprince and Lewis 1983). Sorne species are generalists in the types of areas where they seek hosts, while others are more specialized. Among salt marsh tabanids, for . example, T. nigrivittatus host-seeks around the perimeter of the marsh, Chrysops atlanticus Pechuman prefers open upland areas, and C fuliginosus Wiedemann is most common in the open marsh and adjacent uplands (Dale and Axtell 1976). • In general, horse fly females (Tabanus, Hybomitra spp.) host-seek in open or edge areas, while deer fly females (Chrysops spp.), prefer wooded locations (Golini and Wright 1978). There are, however, many exceptions. Presumably, tabanid females have evolved to select host-seeking sites which maximize their probability of encountering marnmalian hosts. 1.3. Mating Site Selection by Males. The biology of male tabanids is poody lmown, since they are seldom colleeted except from emergence sites (Logothetis and Schwardt 1949, Rockel and Hansens 1970) and from sites where they aggregate in wait of females (Wilkerson ',- • et aL 1985). Pre-mating behaviour by male tabanids usually involves either 74 hovering or perching upon vegetation in a location where females are likely to • pass (Bailey 1948, Wilkerson et aL 1985). Males of most species apparently form leks or mating aggregations, in which males gather in small groups near sorne anomaly in the landscape, such as a hilltop (Miller 1951, Pechurnan 1981, Leprince et aL 1983), a shaft of sunlight filtering through the forest canopy (Blickle 1959), a forest clearing (PechuTJlan et aL 1961, Pechurnan and Burton 1969, Wilson 1967, Pechuman 1981, Magnarelli 1985a), or above the forest canopy (Corbett and Haddow 1962, Philip 1978, Leprince et aL 1983). Mating occurs when females attacted to the anomaly are intercepted by waiting males (Wilkerson et aL 1985). 1.4. Dispersal. • For a short time following emergence, adult tabanids can be found on vegetation near the emergence sites (Miller 1951, Rockel and Hansens 1970). After several hours, their wings have hardened sufficiently to permit flight, and the adults usually disperse; males in search of a lek site and females in search of mating sites and blood meals. Newly emerged female tabanids in the subarctic (Leprince and Maire 1990) and elsewhere (Leprince 1989, Leprince and Lewis 1986) will often seek blood prior to mating. The timing and degree of dispersal by female tabanids is dependent on factors including the female's species and gonotrophic status, the topography and vegetation of the area, and the prevaI1ing • atmospheric conditions. 75 1.4.1. Interspeeific Differences in Dispersal. '. • Females of tabarùd species which are autogenous in their first gonotrophic cycle (i.e., do not require blood to lay their first batch of eggs) generally remain near their emergence sites until they have oviposited; only after they have laid their first egg batch do they disperse and begin active host-seeking (Anderson 1971). Females of anautogenous species (i.e., those which require blood in order to reproduce) on the other hand, disperse from t.heir emergence sites very soon after emergence. Host-seeking female tabarùds ma)' be collected several kilometres from habitats where they emerged (Hanec and Bracken 1964, Bennett and Sr!Ùth 1968, Anderson et aL 1974, Pechuman 1981). 1.4.2. Effeet of Topography on Dispersal. • The amount and type of vegetation in an area affects tabarùd dispersal. Thornhill and Hays (1972) found that the amount of wooded terrain between a release point and recapture traps influenced the number of tabarùd recaptures, and concluded that vegetative barriers interfere with tabarùd dispersal. Roadways are apparently not generally used in dispersal by woodland tabarùds; rather, the flies travel along other unobstructed flight routes through the forest or above the forest canopy (Shepp'\rd et aL 1980). The importance of vegetative barriers to tabanid dispersal has been studied in considerable detail in salt marshes, where there is generally a barrier of dense vegetation between a salt marsh and adjacent uplands. While females of sorne • tabarùd species can readily traverse this barrier and disperse to the uplands, 76 fcmales of other species apparently cannot, and therefore aggregate on the • barricr's marsh-ward side (Rockel and Hansens 1970, Dale and Axtell 1976, Dukes et al. 1974). These species will, however, readily disperse into upland areas through natural or man-made breaks in the barrier (Sheppard et al. 1980). 1.4.3. Erreet of Host Distribution on Dispersal. The distance an host-seeking female tabanid .must travel in search of blood is directly dependent on the proximity of her emergence site to suitable hosts. This, in turn, is dependent upon the local abundance of host types. Tabanids generally feed on medium to large-sized mammals, including cattle, horses, bison, S\vine, deer, caribou, and moose, as weil as smaller hosts such as canids and humans. Sorne tabanid species are more or less host-specifie than others. Female • T. sulcifrons Macquart and H. zonalis (Kirby) are frequently reported biting cattle, but will seldom bite humans, even when cattle are absent (Tasbiro and Schwardt 1949, Smith et al. 1970). Conversely, female H. frontaiis and H. illota (Osten Sacken) will readily bite humans even in the presence of large mammals (Smith et al. 1970, Pechuman 1981). Humans appear to be a preferred host for certain Chrysops spp. (Beesely and Crewe 1963, White et aL 1985). Occasionally there are published accounts of tabanids feeding on small, warm-blooded hosts such as ducks (Davies 1959, Lewis and Bennett 1977), ravens and crows (Davies 1959), osprey nestlings (Jamnback 1969), songbirds (Smith et al. 1970, Lewis and Bennett 1977), rabbits (Magnarelli and Anderson 1980, Lane • and Anderson 1982), and opossums and raccoons (Lane and Anderson 1982), or 7ï cold-blooded hosts such as turtles (Smith et al. 1970). These aCCOUnlS may • represent opportunistic feeding behaviour on the part of hungry flies. but because host preferences are not weil documented for most species (e.g.• by serological determinations of blood meal origins). the possibility of regular blood feeding on avian or other such hosts should not be discounted. In South America. tabanids of the genus Phaeotabanus frequently take blood from reptilian hosts (J.E. Burger. pers. comm). 1.4.4. Erreet ofWeather on Dispersal. AImost ail researchers who have studied tabanids in the field have noted that the activity of the flies, and hence dispersal, is affected in sorne way by environmental conditions. Many authors have noted that tabanid activity does not • occur below a certain threshold temperature, and increases with temperature until an upper threshold is reached (Tashiro and Schwardt 1949, Miller 1951, Blickle 1955, Anderson et al 1974. Dale and Axtell 1975, Lane et al 1983). The lower temperature threshold of tabanid aetivity apparently varies interspecifically (McElligott and Galloway 1991b) and intraspecifically among populations acclimated to different temperature regimes (Baribeau and Maire 1983b, McElligott and Galloway 1991b). Within high and low temperature thresholds tabanid activity may also vary with light intensity (Miller 1951, Anderson 1971, Bumett and Hays 1977, Hollander and Wright 1980, Leprince et al 1983, McElligott and Galloway 1991b), wind speed (Catts and Olkowski 1972, Bumett • and Hays 1974, Dale and Axtell 1975, Lane et aL 1983, Leprince et al 1983), ï8 atmospheric pressure (Bumett and Hays 19ï4, Alverson and Noblet 19ïi, • Kneipert 1982), and relative humidity (Bumett and Hays 19ï4, Dale and Axtell 19ï5. Alverson and Noblet 19ïï, Kniepert 1982). 1.4.5. Dispersal Distance. The dispersal capability of tabanids is potentially very great; Hocking (1953) through f1ight-mill studies, found that females of H. afjinis (Kirby) could f1y up to 91 km on the energy derived from a single crop full of nectar. In the field, tabanids have not been found to disperse nearly as far, although T. [ineo[a Fabricius and T. fuscicostatus Wiedemann in Louisiana forests will travel at least 6.8 km (Sheppard and Wilson 19ï6, Foi! et al 1984). In New York state, Chrysops spp. have been collected from OS to 7 km from their release site after only one • day (White et al 1985). Tropical forest Chrysops have been found to disperse up to 3 km in 6 days (Davey and O'Rourke 1951, Beesely and Crewe 1963) . Generally, however, field data agree with Pechuman et aL's (1983) observation that aithough tabanids have the capacity to f1y great distances, they usually remain fairly close to their breeding sites (Thortthill et al 1971). It should be noted that "as the crow fIies" distances calculated from mark-recapture data may present as inaccurate pieture of aetual tabanid dispersal. White et aL (1985) noted that females of C arer Macquart and C mitis Osten Sacken usually f1ew short distances and then alighted, that repetition of this • behaviour resulted in long-distance travel, and that the direct distance between a 79 release point and recapture point is probably an underestimate of the acmal • distance flown by a tabanid bet\veen the t\vo points. 1.5. Resting Behaviour. During times of inactivity, tabanids rest on such diverse substrates as vegetation, trees, buildings, rocks or bare ground. Inactivity may occur for a variety of reasons, such as the nocturnal cessation of flight activity. weather-related activity suppression, resting following engorgement, or basking for thermoregulatory purposes. The literature on the subject of tabanid resting is mostly anecdotal, except for the study by Kingston et al (1986) on nocturnal resting sites of T. abactor Philip in Texas. At night, this species was found to resl on vegetation, rocks and trees, at heights of 1-1.5 m above ground, and this • nocturnal resting was generally solitary. Engorged females of a number of tabanid species have been observed to rest on foliage for several hours following blood feeding (Clark et al 1976, Sheppard and Wilson 1977, Lancaster and Meisch 1986, Cooksey and Wright 1987). Apparently, tabanids have difficulty flying when fully blood-engorged, and require sorne time to concentrate the blood meal and void exce.;s fluid before they can fly away (Lancaster and Meisch 1986). Thermoregulatory resting has been observed in the beach-dwelling tabanid Apatolestes aetites Philip and Steffan (Lane et al 1983). Flies of this species bask in sun in early morning until sufficiently warmed for flight, and rest on damp • seaweed at midday to cool down. This rather cornplex thermoregulatory 80 behaviour is apparently an adaptation ta life in an exposed (Le., beach) habitat • where temperatures vary from quite cool in early morning to very hot midday. It is likely that other tabanid species also engage in thermoregulatory basking behaviour, particularly in northern areas. Accounts of individuals of several tabanid species resting on roads and paths through wooded areas (e.g., Pechuman et aL 1961) may represent thermoreguhtory basking, but these reports do not specify time of day. 2. GONOTROPHIC DEVELOPMENT OF FEMALE TABANIDS 2.1. Detennination of Follicular Development. The reproductive system of a female tabanid consists of !WO ovaries, an oviduct, a vagina, a pair of spermathecae, and a pair of accessory glands. Each • ovary is enclosed within a membranous sheath, and consists of an internai oviduet into which numerous ovarioles empty radially (Surcouf 1908). In ail Diptera, the ovarioles are of the polytrophic type, meaning that the oocyte and the nurse cells are enclosed within the same follicle. Each ovariole in a reproductively-active female tabanid consists of !Wo follicles. The primary or terminal follicle, in which the developing oocyte is found, is located proximal to the oviduct. The secondary follicle is located distally and the oocyte within it does not develop until the egg developing in the primary follicle has been laid. Terminal (primary) oocytes in both ovaries usually develop • simultaneously, and eggs from the !Wo ovaries are usually laid together as a batch. SI Yolk deposition in the oocytes is a continuous process, but for the sake of • categorization has been divided into five major stages, based upon Christopher's (1911) observations of oocyte development in anopheline mosquitoes. Christophers' Stages I-V have been further subdivided by Mer (1936) and Smith (1970). The foUowing description of the various stages foUows the criteria used by Smith (1970) to gonotrophicaUy age-grade nonhem black flies (Simuliidae), but the stages differ only slightly from those of Mer (1936). At the first, the terminal follicle consislS of a series of 8 undifferentiated ceUs (Stage N), from which the oocyte becomes differentiated (Stage 1). Gradually, very small yolk granules become visible around the oocyte nucleus (Stage IIa). Yolk deposition continues, and eventually the oocyte occupies slightly less than half of the volume of the follicle (Stage lIb). In Stage III, the oocyte • occupies half (ma), then 3/4 (IIIb) of the follicle, which still remains the same size as it was in Stage 1. In subsequent stages of yolk deposition, the oocyte enlarges as yolk deposition continues. By Stage IVa the follicle is 9/10 full of yolk, and has begun to elongate. Stage IVb ovarioles have assumed the shape of the mature egg, and once the chorion has forrned, the oocyte is said to be in Stage V. Stage V oôcyete~ are considered to be mature eggs. Categorization of ovarian development of Tabanidae using Mer's (1936) stages has been carried out en numerous occasions (Raybould 1967, Rockel 1969, Thomas 1973a, Bosler and Hansens 1974, Magnarelli and Pechuman 1974, Magnarelli and Anderson 1977, Magnarelli 1987, Leprince 1989, Lake and Burger 1980, Leprince and Maire • 1990). 82 2.2. Determination of Parity. • Gonotrophic age-grading of tabanids, often termed physiological or reproductive age-grading, is based upon the fact that it is possible to differentiate females which have laid eggs (parous, P) from those which have not (nulliparous, N). The most obvious physical evidence that a fly is parous is the presence of unlaid eggs in the reproductive system. Typically, however, the proportion of parous flies which retain eggs is very low, e.g., 2-14% (Gordon et al 1948, Lewis 1960, Magnarelli and Pechuman 1974, Thompson et al 1979a,b, Auroi 1982, Leprince and Lewis 1986, Leprince and Bigras-Poulin 1990), and therefore this method is of little use in consistently determining parity. The most commonly-used and reliable method of establishing parity is based upon examination of the membranous follicular tunica, and was fust • demonstrated by Polovodova using anopheline mosquitoes (Detinova 1962). In the ovaries of nulliparous individuals, the tunica between each developing ovariole and the central oviduct forms a clear tube or pedicel of relatively even diameter. Immediately following oviposition, the tunica beyond the secondary follicle is distended, sac-like and about the size of the recently-shed egg (Thomas 1972). Over a perlod of approximately 48 hr, this sac contraets and the cellular debris within it becomes concentrated into a follicular relie. Parous flies with ovarioles up to and including this stage are termed parous saccuIate (psac). Further contraction of this dilatation continues until a discrete oval or spherical dilatation remains. Flies with ovarioles in this stage are termed parous yellow-bodied (Pyb) • (Thomas 1972), because the follicular debris is often yellow in colour. In mosquitoes, it is possible to identify the number of gonotrophic cycles an • individual has completed based upon the number of yellow bodies in ilS pedicels, since an individual forms one body after each oviposition. ln tabanids, however, il is usually only possible to tell whether or not an individual is parous, since yellow bodies from multiple gonotrophic cycles may combine into a single dilatation (Magnarelli and Stoffolano 1980). Sorne authors, however, have collected individuals which displayed !Wo or three yellow bodies in their pedicels, and had therefore successfully completed !Wo or three gonotrophic cycles (Thomas 1971, 1973a, Thompson et al 1979b, Lane and Anderson 1982). 2.3. Autogeny and Anautogeny. Blood feeding in female tabanids is doubtless their best-known trait; that • horse flies and deer fIies are blood-feeders can be attested to by anybody who has spent a surnrner afternoon in an area where they are common. Blood feeding is the basis of the importance of tabanids as pests of humans and livestock, and as disease vectors (e.g., Twinn et aL 1948, Chvala et al 1972, Krinsky 1976, Perich et al 1986, Foil 1989). In aIl tabanid species which blood-feed, blood is taken strictly as a means of providing nutrients for developing oôcytes; energetic needs are met through feeding on nectar and other sugar sources (Wilson 1967, Magnarelli and Anderson 1977, Magnarelli et aL 1979, Magnarelli 1987). Fernale tabanids are either autogenous, i.e., not requiring a blood meal in order to produce eggs, or anautogenous, i.e., requiring blood for their fust ovarian • cycle. In the case of autogenous females, fat-body accumulated during the larval 84 stages is sufficient ta permit develapment of a first egg batch without the • requirement for nutrients from an external source (Thomas 1973a, Bosler and Hanscns 1974, Leprince and Lewis 1983, 1986, Magnarelli 1985b,c), whereas anautogcnaus females either lack sufficient larval fat-body to produce eggs autogenously, or are physiologically incapable of doing so without a blood meal (Morris and Defoliart 1971, Troubridge and Davies 1975, Leprince and Lewis 1986). 2.3.1. Autogenous Species An autogenous first ovarian cycle is apparently the rule for species such as the salt marsh tabanids C atlanticus, C fuliginosus, and T. nigrovittalUS (Rockel 1969, Anderson 1971, Bosler and Hansens 1974, Magnarelli and Stoffolano 1980), • and numerous other tabanid species (Lake and Burger 1980, Magnarelli 1987, Leprince 1989). In field studies, a species is considered to be autogenous if no nulliparous host-seeking individuals are colleeted (Thomas 1972). In species such as C univittatus Macquart, certain populations are apparently autogenous while others are anautogenous (Leprince and Lewis 1983, Magnarelli and Anderson 1981). In still other species, autogenous and anautogenous females are suspeeted to occur in the same population (Thomas 1971, Auroi 1982, Troubridge and Davies 1975, McElligott and Galloway 1991a). This latter case, termed facultative autogeny, may occur if autogeny is ooly expressed in those females which emerge from pupation with fat-body sufficient for oocyte maturation (Lake and Burger 1980). Fat body carried through • -- pupation. in turn, may be influenced by larval habitat qu:\lity; well-fed 1:\I'\'a.: are • thought lO become autogenous adults, whereas poorly fed larvae become anautogenous ones (Lake and Burger 1980). This is the case with facultatively autogenous mosquitoes (Spielman 1957, 1971. Q'Meara and Craig 1969). Differences among populations in autogeny may reflect differences in larval habitat quaiity between areas. just as autogenous and anautogenous individuals within the same population may represent flies emerging from portions of a habitat of different degrees of quality. Alte:natively, differences between populations and bet\veen individuals within populations may reflect strictly . physiological (i.e., genetic) differences. In females of most autogenous species, eggs are only produced autogenously for the first ovarian cycle. Following this, autogenous femaies • essentiaily become anautogenous, and require blood to produce subsequent egg batches (Rockel 1969, Thomas 1973a, Magnarelli and Stoffolano 1980, Magnarelli and Anderson 1981). In rare instances tabanid species are entirely autogenous and are able to complete multiple gonotrophic cycles without ever having to blood feed. Apatolestes actites for example, never blood feeds and consequently the mouthparts of femaies are reduced (Lane and Anderson 1983, Lane et al 1983). 2.3.2. Anautogenous Species. The majority of North American tabanid species appear to be obligately anautogenous, in that their ovarian follicles will develop no further than • Christopher's stage Il unless the femaie blood-feeds (Thomas 1973a, Magnarelli 86 and Pechuman 1975, Lake and Burger 1980, Lane and Anderson 1982, Leprince el • al. 1989, Leprince 1990). Anautogeny in these species evidently has a genetic basis, and occurs consistently throughout the broad geographic ranges of such spccies as H. lasiophlhalma (Morris and Defoliart 1971. Thomas 1972, Troubridge and Davies 1975, Magnarelli 1976, Thompson el aL 1979b, Leprince and Bigras Poulin 1990) and T. quinquevillalus Wiedemann (Magnarelli and Pechuman 1975, Troubridge and Davies J975, Magnarelli 1976, Leprince and Lewis 1986). 2.4. Timing or Ovarian Dcvclopment in Tabanids. For sorne tabanid species (e.g., anautogenous Chrysops spp., autogenous and anautogenous Tabanus spp.), yolk deposition begins only after the female has emerged from the pupal stage. Newly emerged females of the aUlogenous tabanid • T. nigrovillalus invariably have NI ovarioles for the first day post-emergence, but have developed NIII-IV ovarioles by 4 days post-emergence (Magnarelli and Anderson 1977). Similarly, for the anautogenous species C mitis, only one day old females had NI follides; thereafter, only NIl follides were present in females which had not yet blood fed. Unfed anautogenous female C excitans Walker, C indus Osten Sacken, C. shennani Hine, and C vit/atus Wiedemann all had developed ovarioles to stage II by 3 days post-emergence (Lake and Burger 1980) In autogenous Chrysops spp., however, the timing of ovarian development in non blood-fed flies is very different. Yolk deposition begins in the ovarioles of a female C fuliginosus while she is still in the pupal stage. By the time the female • emerges, her ovarioles are in Stages II-IV, and oviposition occurs after only 7-10 .'-~., ,1 days (Magnarelli and Anderson 19ï7). The same is true for aUlogenous female C. • arer; day-old females had NIl-III ovarioles, and many females had developed mature eggs by 7 dàys post-emergence (Lake and Burger 1980). Due to the difficulty in inducing tabanids to blood-feed, mature eggs, and oviposit in the laboratory (Roberts 1966. Thompson et al. 1979, 1980), !iule is knovm of the timing of ovarian development in parous flies. lt is known, however, that the Psac stage is indicative of recent oviposition; it persislS for 48 hours or less post-oviposition (Thomas 1972). 3. TABANID RE5EARCH IN NORTHERN CANADA. 3.1. The Tabanid Fauna of Northern Canada. Teskey (1991) recently published a handbook of Canadian Tabanidae, in • which the knl)wn geographic ranges of the 145 tabanid species found in Canada were summarized. Approximately 40 tabanid species are found in boreal and suharctic regions of Canada. Most (23 species) are Hybomitra and Chrysops spp. which have northem transcontinental distributions (the bulk of their geographic ranges are located in boreal and subarctic Canada), while the ranges of ooly a few typically southem species (8) penetrate northward into boreal regions. The remainder of species are restricted geographically to either the east or west of Hudson Bay. The northern !imit of the range of most tabanid species in northern Canada is marked by the tree !ine, although species such as H. frontalis and H. • sexfasciata (Hine) have been colleeted on the aretic tundra. 88 Seventy five percent of northern tabanid species (30/40) are common at • lcast in sorne part of their geographic ranges, and often a number of species will be common in a given area. The result is that many areas of the north are blcssed with an :. Jundant, if not tremendously diverse, tabanid fauna during the short summer season. 3.2. Seasonal Activity of Northern Tabanids. Throughout Canada, adult horse flies and deer flies are active only during the warm months of the year, and there is typically a marked succession of species present over the course of the summer months. The seasonal activity of tabanids in southern Canada usually extends from mid-May until early September, and seasonal species successions have been described in the Maritimes (Lewis and • Bennett 1977), Québec (Lewis and Leprince 1981, Leprince anà Lewis 1983, Thibault and Harper 1987), Ontario (Davies 1959, Pechuman et aL 1961, Smith et aL 1970, Golini and Wright 1978), and Manitoba (Hanec and Bracken 1964, McElIigott and Galloway 1991a). As summer length varies latitudinally, so too does the length of the tabanid flight season; in northem (e.g., boreal and subarctic) regions, tabanid flight activity is largely confined to the months ofJuly and August or parts thereof (Twinn et aL 1948, Miller 1951, Maire 1984). 3.3. Other Research on Tabanids in Northern Canada Apart from strictly faunistic accounts (e.g., Freeman 1953, Weber 1950, • Judd 1967), relatively little has been published conceming the biology of tabanids in nonhem Canada. Hanec and Bracken (1964) and Thomas (1973b) examined • the geographical distributions of tabanids in borea! regions of Manitoba and Alberta, respectively. Simii:lrly, Baribeau and Maire (1983b) and Maire (1984 a,b) described the distribution of tabanids in subarctic regions of northern Québec. Twinn et aL (1948) and Miller (1951) reported the flight seasons of various tabanid species near Churchill, Manitoba, and Baribeau and Maire (1983b) and Maire (1984 a,b) also presented flight seasons of tabanids in northern Québec. Miller's (1951) study still provides us with the only detailed account of tabanid biology in subarctic Canada; his work at Churchill has provided important insights of tabanid seasonality, relaùve abundance, emergence, nectar-feeding, maùng, climaùc effects on tabanid activity, and larval bionomics in the north. • James (1953) followed up Miller's study with an account of the predators and parasites of larval and adult tabanids in the Churchill area. Maire and Beaudoin (1984) examined specimens of H. zonalis (Kirby) and H. aequetincta (Becker) from various locations in northern Québec, in order to taxonornically resolve the !wo species. McElligott and Galloway (1991b) compared the daily activity patterns and climatic influences on the daily activity of !wo Hybomitra spp. at sites in the southem and northem Manitoba, as Baribeau and Maire (1983a) compared tabanid faunas and periods of flight aetivity be!Ween sites in temperate and subaretic Québec. Hocking (1952), in a general treatment of northem biùng flies, pointed out that tabanids, together with mosquitoes and • blâck flies, have both physiological (i.e., pain, blood 1055, disease transmission) and 90 psychological (Le., constant humming in the ears, inability to escape flies, panic) • effects on their human hosts. Leprince and Maire (1990) dissected specimens of Hybomitra spp. collected at Richmond Gulf in northern Québec, to determine their parity, stage of yolk deposition, and sperm presence. This is the only published account of tabanid gonotrophic age-grading carried out in northem Canada. • • 91 • II. INTRODUCTION In subarctic Canada adult tabanids are active only during the few brief weeks of surnrner. Throughout this period the blood-feeding activities of the female horse flies and deer flies can cause considerable annoyance to man and other marnrnals (Miller 1951, Hocking 1952). Female tabanids may also serve as vectors for parasites and pathogens (KrinsJ...-y 1976, Foil 1989). Regional information concerning the seasonal population dynarnics of tabanids is important because a knowlege of seasonal abundance patterns, . combined with information concerning the annoyance or vector potential of each species, allows identification and anticipation of the periods when these insects present the greatest nuisance or health risk, and when steps should be taken to • minimize their impact. The seasonal activity of tabanids in the subarctic is poody known relative to more southem regions of Canada. In general, northern tabanid studies have been carried out opportunistically e.g., a research tearn working on mosquitoes at a given locale has collected tabanids as weil. Consequently, detailed descriptions of tabanid abundance in the subaretic are few. The periods of flight aetivity and/or pattertts of seasonal abundance of sorne tabanid species in subarctic Québec and Labrador have been described in varying degrees of detail by Baribeau and Maire (1983b), Maire (1984a,b), and Leprince and Maire (1990). In northem Manitoba, Miller (1951) described the seasonal occurrence and • relative abundance of common tabanid species. One objective of the present 92 study was to provide a detailed account of the relative abundance and seasonal • occurrence of host-seeking females of tabanid species near Schefferville, Québec. A second objective was to compare the relative abundance of tabanids between two different peatlands, to determine the degree to which tabanid catches vary locally. Only Leprince and Maire's (1990) study has previously exarnined parity and gonotrophic age of subarctic tabanid populations. A third objective of my study was to deterrnine seasonal changes in parity and gonotrophic age of the female tabanids, to ascertain which species at Schefferville were autogenous or anautogenous, and to determine the timing of adult emergence. Finally, gonotrophic age-grading data from the two peatlands were compared, to find out • if seasonal emergence patterns differed among the sites. III. MATERIALS AND METMODS 1. Study Areas During the summers of 1990 and 1991, adult tabanids were trapped on a daily basis at two peatlands near Schefferville, Québec. The fust site, Iron Arm fen, is located in Labrador 20 km ne of Schefferville (Fig. 1). This fen is essentially a large (580 x 120 m), weI, open meadow; its predominant vegetation is sedges (Carex, Scïrpus spp.) and the shrub Myrica gale Linneus, with cotton grass • (Eriop/zorum spp.) and mosses (e.g., Sphagmun spp.) common in the wetter areas. 93 The fen is surrounded on ail sides by rel:ltively open lichen-spruce woodland, • which gives way to denser spruce forest approximmely 100 m from the fen margin. The second sampling location, Capricorn fen, is a relatively small (160 x 70 m) peatland, located approximately 2 km south of Schefferville (Fig. 1). This fen, apart from size, is similar to Iron Arro fen in its general appearance and vegetation. 2. Collection of Adult Tabanids, Methods of collecting adult tabanids are reviewed by Teskey (1990); canopy traps (Catts 1970) and Malaise traps (Townes 1972) were used ta sample populations of adult tabanids during the present study. Unbaited canopy traps utilize highly specifie visual eues ta attract host-seeking female tabanids, and • therefore collect little else. While they are highly effective for collecting most Hybomitra spp., canopy traps are generally less effective for catching Chrysops spp., and are almost useless for trapping Atylorus spp. (Pechuman and Burton 1969, Roberts 1976). The attractiveness of canopy traps, and the variety of species they collect, can be enhanced by the release of carbon dioxide gas near the trap (Roberts 1971, 1975, McElligott and McIver 1987), however this was impractical at Schefferville owing ta the isolation of the field sites. In an attempt ta collect as many as possible of the tabanid species present, Malaise traps were used in addition to canopy traps. A Malaise trap is a non attractive fligbt intercept trap which colleets flying insects indiscrirninately (Martin • 1977). While unbaited Malaise traps do not collect host-seeking females in as 94 great numbers as canopy traps, they collect ail types of flying tabanids, regardless • of whether the flies are host-seeking, or whether or not they are attracted to visual targets (Roberts 1976). One canopy and one malaise trap were located in the ecotone at the edge of each peatland; tabanid host-seeking activity is apparently most concentrated in edge areas (Golini and Wright 1978). Trapping was carried out continuously in 1990 for 47 days from 25 June to 20 August. A sampling interval began when traps were emptied at approximately 1630 hr EDT one day and ended when traps were emptied at 1630 hr the following day. In 1991 sampling was conducted for 46 days from 24 June until 20 August. Trapping periods in 1990 and 1991 as far as was known encompassed the flight seasons of all tabanid species present at both sites. • 3. Categorization of Ovarian Development. Whenever the daily catch of a tabanid species met or exceeded 10, a sample of flies was frozen and retained for later dissection and gonolrophic age grading. In ail, 1-5 daily samples were collected and disseeted for each week of a· species' observed flight season (Appendix 1). From each of these samples, the percentage of flies in each stage of ovarian development was calculated. The average weekly percentages of flies in each of the different stages of ovarian development were then calculated and plotted for cach week·of the flight season. • 95 3.1. Dissection. • Gonotrophic age was determined for 10 flies from each ~ample by e.xaminaùon of ovarioles. Ovaries were e.xposed by removing th.:: two terminal abdominal segments and opening the abdomen along the right pleural membrane. This technique allo\Ved easy removal of complete ovaries and also permitted exarnination of the abdominal cavity for the presence of fat body. Ovaries \Vere removed from the abdomen using #5 \Vatchmaker's forceps, placed on a slide in :\ drop of disùlled water, gently teased apart, and the individual ovarioles vie\Ved against a black background using a dissecting microscope (120-250X). 3.2. Determination of Parity. Female tabanids \Vere recorded as nulliparous (never having laid eggs), or • parous (having laid at least one egg batch) based on their state of ovarian development (see Thomas 19713). No attempt \Vas made to separate multiparous from uniparous flies, since multiple gonotrophic cycles in tabanids do not always result in multiple follicular dilatations (Magnarelli and Stoffolano 1980). 3.3. Gonotrophic Age-Grading. The stage of yolk deposition in the ovarioles of both parous and nulliparous flies \Vas recorded as Stage I-IV, according Smith's (1970) modification of Mers (1936) stages. Several assumptions based upon previous ovarian age • grading studies were made during the present study: 96 a) If a female is obligately anautogenous, yolk deposition proceeds no further • than Stage II in the absence of a blood meal. Tabanids of obligately or facultatively autogenous species will accumulate yolk beyond Stage II without a blood meal only during the first ovarian cycle; parous females of these species behave as anautogenous flies. b) Anautogenous flies with l1ulliparous Stage 1 ovarioles are recenùy emerged (i.e., within 24 hr of emergence (Magnarelli and Anderson 1977, Lake and Burger 1980). c) Parous sacculate females have ovipositerl within 48 hr capture (Thomas 1972). 4. Categorization of Fat Body Deposition. As weil as gonotrophic age, the amount of fat body present in the abdomen • of each fly dissected was recorded on a scale of 0 to 3, as follow~: 0: no fat body visible in the abdomen under 10x25 magnification; 1: small amount of fat visible, forming a thin semi-transparent layer over the inside of abdominal tergites and stemites; 2: fat layer on inside of abdominal segments thicker, opaque; 3: thick, convoluted layer of fat present in abdomen. Although this categorization was somewhat subjective, it was applied consistenùy by the author only. • 97 For ail flies dissected of a given species. the number of flics in cach • gonotrophic age category was ca\culated. The percent of flies in each category which had a given stage of fat deposition was then ea\culated and plotteù. III. RESULTS AND DISCUSSION 1. Faunal Diversity. 1.1. Faunal Diversity in the SchelTerville Area vs. Other Areas. Total nurnbers of host-seeking female tabanids of each species collecied at Iron Arm and Capricom fens, as weil as the duration of the observed flight seasons in 1990 and 1991, are presented in Table 1. In ail, 29,736 tabanids in 3 genera and 17 species were collected in the Schefferville area. Horse flies • (Hybomitra, Atylotus spp.) comprised nearly 96% of ail tabanids collected; the remaining 4% were deer flies (C/zrysops spp.). 1.1.1. Tabaninae. Hybomitra arpadi (Szilady) was the most abundant horse fly in the trap collections in the Schefferville area, comprising 54% of the total tabanids trapped (Table 1). This species is common elsewhere in nonhem Québec; it made up 42% of the total tabanid catch at Lake Delorme (55°00'N 700 00'W), 160 km west of Schefferville (Banbeau and Maire 1983b), and 35% of the tabanids collected at Richmond Gulf on Hudson Bay (56"20'N, 75°80'W) (Maire 1984b). At another • location on Hudson Bay, Poste-de-la-Baleine (55°17'N 77"45'W), however, H. 98 arpadi was ahsent from collections (Maire 1984a). At Churchill, Manitoba. in the • western subarctic, this species (as T. gracilipalpus) was collected commonly by Miller (1951). Next in abundance to H. arpadi at Schefferville was H. aequetincta, which made up 31 % of the total tabanid catch (Table 1). This species is apparently more common at Schefferville than it i~ in sorne other areas of nonhern Québec: it comprised 12% of the tabanids collected at Richmond Gulf, and less than 4% of those collected at Lake Delorme. Like H. arpadi, this species was not collected at Poste-de-la-Baleine. Other species collected commonly at Schefferville were Hybomitra zonalis, H.lurida (Fallén), and H. Izearlei (Philip) (Table 1). Hybomitra zonalis made up a considerably smaller percentage of the tabanid fauna at Schefferville (7%) than it • did at Lake Delorme, where over 14% of the tabanids colleeted were of this species. At Richmond Gulf, however, H. zonalis made up from 0-14% oÏ the total tabanid catch (depending on year), and at Poste-de-la-Baleine, this species was very rare. Hybomitra lurida made up 2% of the tabanids collected at Schefferville, and made up approximately the same proponion of the tabanid catch at Lake Delorme and Richmond Gulf. At Poste-de-la-Baleine, however, this species was one of the most commonly collected species, making up nearly 30% of the total flies collected. Hybomitra Izearlei constituted a much smaller percentage of the tabanid • fauna at Schefferville (2%), than il did at Richmond Gulf (25-30%). This species was apparently rare at Poste-de-la-Baleine. and ab"cnl twm <:l,lle,tilllb al ~Ik,' • Delorme. Three horse fi) species, H. astwa (Osten S:\cken), li. jroll/alis. and H. pec/lumani Teskey & Thomas. were collected consistently throughout their observed flight seasons al Schefferville. but never very commonly: each represented less than 1% of the total horse fly catch (Table 1). Hybolllitra aswttl is considered a rare species elsewhere in northern Québec (Baribeau and Maire 1983b, Maire 1984b). and is apparently uncommon throughout its transcontinental range (Teskey 1991). Hybomitra pec/zul1lani. on the other hand. is a corn mon species in southern Canada (McElligott and Galloway 1991a. Baribeau and Maire 1983a), but is only distributed sporadically in northern Québec. This species was apparently not collected at Lake Delorme, Richmond Gulf, or Poste-de-la • Baleine, although il has been take:: previously from Schefferville, from Port Harrison (58"27'N 76°06'W), and from Fort George (53°50'N 79°00'W) on the eastern shore of Hudson Bay (Teskey 1991). Hybomitra frontalis, the third uncommon species at Schefferville, was rare in collections at Lake Delorme, but common in collections at Richmond Gulf. This species was the most abundant tabanid at Poste-de-Ia-Baleine, where it made up 84% of the total tabanid catch. At Churchill, Manitoba, H. frontalis is the also the most common tabanid species (Miller 1951, McElligott and Galloway 1991b). The remaining three horse fly species collected at Schefferville, H. a/finis, H. liorhina (Philip), and Atylotus splzagnicola Teskey, were anly rarely callected • thraughaut their f1ight seasans at Schefferville, and eaeh accaunted far less than 100 0.1% of the IOtal horse flv catch (Table 1). While H. affinïs is a common species • in lhe weSlern subarclic (Miller 1951, McElligott and Galloway 1991b), il has scldom been collecled in Québec and Labrador (Teskey 1991). Maire (1984a,b) and Baribeau and Maire (1983b) apparently did nOI collecl H. affinïs during their sludies, allhough they may have rnistaken H. arpadi for this species. Hybomitra liorhina, and A. sphagnicola, also rarely collecled at Schefferville, are rare species throughout their geographic ranges (Teskey 1991). Maire (1984) colleCled only two specimens H. lior/Zina at Richmond Gulf, and this species was not taken al Lake Delorme or Poste-de-la-Baleine. Other than at Richmond Gulf, H. lior/zina has only previously been collected in northern Québec-Labrador at Fort Chimo (5~006'N 64"24'W) and Schefferville (Teskey 1991). Sirnilarly, A. sp/zagnicola has previously only been reported from three locations in northern • Québec·Labrador: Fort George, Lake Mistassini (51°00'N 73°37'W). and Camvright (53°42'N 57"Ol'W). 1.1.2. Chrysopsinae. Chrysopinae were collected much less commonly than Tabaninae at Schefferville; only 1,232 of the 29,736 tabanids colleeted (4.1%) were deer flies. • Chl): tOlal C/zrysops catch and 2.5% of the total tabanids (Table 1). The remaining Clzl)'sopS species each made up less than 1% of the total tabanids. Chrysops nigripes Zetterstedt made up 17% of the C/zrysops colleeted, followed by C ater • \vith 9% and C furcarus Walker \vith 7%. l.ess common species were C frigidus llli OSlen Sacken (2.5%), C. excirans (1.5%). and C. sordidllS OSlen Sackcn (1.2':,,) • (Table 1). The abundance of C. :in::aills al Schefferville is surprising, since lhis species is apparently uncommon in Canada (Teskey 1991). Baribeau and Maire (1983b) and Maire (1984a,b) apparently did not colleCl lhis species in norlhern Québec, although il is easily confused \Vilh C. Iligripes. especially in keys pre daling lhat of Teskey (1990). Chrysops nigripes, C. arer, and C. frigidllS have been collected regularly. lhough relaùve1y uncommonly, from other areas of norlhern Québec (Baribeau and Maire 1983b, Maire 1984a,b). Chrysop! furcatus \Vas apparently rare at Richmond Gulf, but was one of lhe most common tabanid species at Lake Delorme and Posle-de-Ia-Baleine, making up 16-23% of the total tabanids • collected there. Similarly, C. excitans, while common al Lake Delorme and Poste de-la-Baleine, was not collecled at Richmond Gulf. The opposite was true for C. sordidllS; this species was collecled commonly at Richmond Gulf, but was rarely collected at Posle-de-la-Baleine, and absent from collections at Lake Delorme. 1.2. Local Faunal Diversity: Inter-Annual and Inter-Site Variation. 1.2.1. Tabaninae. The number of flies collected varied considerably among sites for certain tabanid species at Schefferville. In both 1990 and 1991, more than three times as many H. arpadi were taken at Iron Arm fen as at Capricom fen (Table 1). • Similarly, the number of H. aequetincta collected at Iron Arm in 1990 was much 10~ grcalcr lhan al Capricorn. In 1991, however. more H. aequetincta \Vere taken at • Capricorn than at Iron Arm. At Iron Arm, the catch of H. arpadi in 1990 \Vas nearly t\Viee that eoIIeeted in 1991 (Table 1). In 1991, catches of H. aequetincra \Vere much higher at both Iron Arm and Capricorn than in 1990. Varying degrees of inter-site and inter annual variation \Vere also apparent in numbers of H. fronralis, H. hearlei, H. lurida. H. peclzumani and H. zonalis (Table 1). 1.2.2. Chrysopinsae Arnong Chrysops spp., variation between years and sites was aIso apparent in numbers of flies eolleeted. At Iron Arm, C. furcarus \Vas eoIIected relaùvely eommonly in 1990, but \Vas rare there in 1991. At Capricorn, this species was • practically absent in both 1990 and 1991 (Table 1). Sb:ty five C nigripes \Vere eollected at Iron Ann in 1990, ve= ooly 21 in 1991. Conversely, only 42 C. nigripes \Vere coIIected at Capricorn in 1990, but 81 \Vere collected in 1991 (Table 1). Approximatcly the samc number of C. zinzalus were collected at Iron Ann in 1990 and 1991, but the catch of this species at Caprieorn in 1991 was nearly t\Vice that taken in 1990 (Table 1). In 1990, but not in 1991, the catch of C. zinzalus \Vas notably higher at Iron Ann than at Capricorn. Numbers of C. ater and C. frigidus aIso varied among sites and years. • 1.2.3. Uncommon and Rare Species. • Sorne tabanid species (e.g.. H. a/finis. H. mtllta. .-1. spl/(/gllicola. C. ,:rCÎ(IIl/s). were collected uncommonly during both years of study ;\l Iron Arm ;md Capricorn, but other species (e.g., H. iiorizina. C. sordidus) were absent l'rom collections during certain years at either Iron Arm or C;lpricorn (Table 1). 1.3. Factors AlTecting Faunal Diversity: General Discussion. . 1.3.1. Geographie Variation in Faunal Diversity. It is very difficult to characterize the tabanid fauna of a given area with any degree of certainty. In northern Québec and Labrador there exists tremendous variation among geographically similar locales, both in the density and the diversity of tabanids captured, as was detailed in Sections 1.1.1-1.1.2. Bcsides • faunal variation on a broad geographic scale, smaller scale variation, such as that apparent between the faunas of Iron Arm and Capricorn fens, further compound the problerns of obtaining an accurate survey of an area's tabanids. Localized variation in diversity and abundance of tabanids has aiso been noted elsewhere, among collection sites in bog and forest at Richmond Gulf (Maire 1984b), among peatlands near Trois-Rivières, Québec (Bariheau and Maire 1933a), and among collections made in various habitat types in southwestern Ontario (Golini and Wright 1978). Geographie variation in tabanid numbers can be attributed to a number of factors. On a broad scale, the climatic tolerances of tabanids determine where • they can or cannot exist. North of the tree line, the generai absence of tabanids 104 has becn attributed ta the faet that summer thaw is of too shon a duration 10 • permit tabanid larval development to oeeur (Maire 1984a). Temperature ean also affect adult tab:mids directly; thresholds for adult tabanid activity vary intcrspeeificaIly, and interspecificaIly among different latitudes (Baribeau and Maire 1983, McEIligott and GaIloway 1991b). At high latitudes or high altitudes, where temperature is generaIly too cool to permit flight activity, tabanids are coIlected only rarely or not at aIl (Maire 1984a). Geographie variation in availability of larva! habitats, on both a broad scale and a local scale, also appears tO be important in determining tabanid distribution and local abundance. On a broad scale, the relationship between larval habitat and adult abundance is especially notable for species with narrow larva! habitat requirements, as noted in the Literature Review section of this chapter. For • peatland-breeding tabanid species, the availability of bogs and fens can have a profound effect on adult abundance (Maire and Beaudoin 1984, Maire 1984b, Lewis 1987). Local variation in the availability of suitable breeding habitat will also affect tabanid numbers. In the present study, C. furcatus was much more common at Iron Arm than at Capricorn, presumably because Iron Arrn was more suitable than Capricorn as larval habitat. Locally, the interacting factors of the host-seeking-site preferences of host-seeking fema!e tabanids, vegetation, and the general landscape of an area are also important in determining tabanid distribution (see Literature Review section). Fina!ly, on both broad and local • geographic scales, the availability of large mammalian hosts is thought to IO~ influence tabanid abundancc (Maire and Beaudoin 1984, T.D. Gallllway peTS. • comm.). Apparent differences in the tabanid fatmas of different areas C;tn, in some instances, be tmced ta differences among studies in methods uscd to trap tabanids. Canopy traps, for example. will not collect certain tabanid spccies; if only one type of trap is used, a biased picture of a local t;lbanid f;\llna may be obtained. During the present study. canopy traps \Vere completel)' incffectivc in collecting A. splzagnicola. even thOllgh they collected HyiJomirrcl and Chrysops spp.: A. splzagnicola was only taken in Malaise traps. 1.3.2. Annual Variation in Faunal Diversity• As weil as varying geographically, a tabanid fauna can vary annllally. Sorne • tabanid species may be collected consistent!y at a given locale, but vary dramatically from year to year in abllndance. Other, rarer species may be present in trap catches one year but absent the next (Maire 1984a,b, McElligotl and Galloway 1991a). Annual variation in adult tabanid numbers is most likely a consequence of localized year ta year variation in size larval cohons, as outlined for C. zinzalus and C. nigripes in the preceeding chapter. Inter-annual variation in extrinsic factors, sllch as local availability of hOSL~ and the number of days with climatic conditions suitable for host-seeking, also likely affect local tabanid abundance. In a year when hosts are common in an area, tabanids from that area are most likely ta remain there. Conversely, during • year when there are few hosts in area "A", females from there are more likely to 10(, diffus\; intel arca "il" to find a blood meal, thus affecting the tabanid abundance in • arca "B". If warm sunny days prcdominatc during a givcn summer, female tabanids host-scck morc and potcmially disperse further than during a summer dominated by cool, overcast days. In a warm summer, females are therefore more Iikely ta occur in a given area, even if they did not emerge locally, and are more likely ta hc collectcd. In a cool summer, however, tabanids may not be collected commonly, simply because their host-seeking activity is curtailed or restricted (Maire 1984b). 2. Tabanid Flight Seasons in the Schefferville Arca• 2.1. Host-Secking Females. • Inclusive flight seasons of host-seeking tabanids at Iron Arm and Capricorn extended from 25 June ta 16 August in 1990, and from 4 July until 11 August in 1991 (Table 1). Flight season duration was remarkably consistent between sites and years of study; dates of first and last appearance of a given species were similar in 1990 and 1991, and at Iron Arm and Capricom fens, except for rarely collected species (e.g., C. exdtans), where capture of a specimen occurred sporadically, and accurate estimation of flight season length was not possible (Table 1). Tabanid flight seasons observed at Schefferville were aIso consistent \Vith the findings of other studies in northern Québec (Baribeau and Maire 1983a,b, Maire 1984a,b,), except for minor variation attributable ta Iatitudinal • variation in the onset of summer. There is a definite seasonal succession of species amol1!:-the host-seekil1!:- • female tabanids in the Schefferville area. HyIJolllirf(l Il/rida. fi. (/t'ql/cri/l(;ra. H. arpadi, and H. zonalis are the first species ta appear, and :111 make their appearance more or less simultaneously, either in the 1:Ist week of June or the first week of luly (Table 1). Hybomirf(l Il/rida is common only in carly July, but :1 few individuals persist into August (Figs. 2,3). Hybomirra acqucrincta, fi. arpadi and H. zonalis are present throughout the entire tabanid tlight se:lSon, :llthough numbers vary considerably depending on when peak activity occurs (Figs. 2,3) . Later in the summer, H. hearlei, H. froncalis, and H. asrwa make their appearance (Table 1). They are active l'rom mid-luly umil early August (Figs. 2,3), and this is also generally true for Chrysops spp., although species such as C. excitans, C. nigripes, C. arer, and C. zinzalus may appear in early luly in sorne years (Figs. 4,5). • The last tabanid species to emerge in the Schefferville area are apparently H. pechumanl and H. Ilorhina, which are active only l'rom the last week of luly and early August (Figs. 2,3). Hybomitra Ilorhlna is one of the last tabanids to appear in western montane parts of its range (Burger, pers. comm.). Within the observed tlight seasons of the more common tabanid species. seasonal peaks of activity were observed (Figs. 2-5). Peaks occurred either at the onset of the tlight season, or at approximately the midpoint of the season. For H. lurida, peak numbers of host-seeking females were always encountered within a few days of the beginning of tlight activity, whereas peak populations of H. zona/Is, H. hear!ei, H. pec/zumanl (Figs. 2,3), and ail C/zrysops spp. (Figs. 4,5) occurred at • or near the middle of the tlight season. lOS The seasonal peak of host-seeking activity did not oecur eonsistent!y at one • point in the flight seasons of H. aequerillcra and H. arpadi, but rather varied between sites and years of stuùy (Figs. 2,3) At Capricorn, peak numbers of H. llequerinc/(l were trapped early in the flight season in 1990 anù 1991 (Fig. 3). At Iron Arm, however, the peak of this species was approximately midway through the flight season in 1990, but was al the beginning of the flight season in 1991 (Fig. 2). Similarly, at Capricorn, peak catches of H. arpadi in 1990 were made in miù-season, while in 1991, peak numbers were trapped at the beginning of the flight season (Fig. 3). Such marked inter-site and inter-annual variation in the timing of seasonal activity peaks has not, to my knowledge, been reported previously in the literalure. In Figures 2-5, considerable day to day fluctuation is apparent in the • numbers of tabanids of each species collected. This is mainly a consequence of daily variation in the weather. since certain cIimatic conditions suppress or prevent tabanid flight activity. This is discussed in detail in the next chapter. 2.2. Males. Male tabanids were rare in canopy and Malaise trap collections; only 72 individuals were collected. ln general, male tabanids are not k-nown to linger at emergence sites, but migrate ta a hilltop or other lek site soon after emergence (Wilkerson et aL 1985). Those specimens trapped at the fen were aImost certainly recently-emerged individuals which followed females into the traps. Forty eight • male H. arpadi were taken, 35 from 8-27 luly 1990, and 14 from 5-30 luly 1991. !lN Eleven male H. zonalL,' were uapped in the period l'rom 14-1~ Jllly jlNO; no maks • of this species were collected in 1991. Three males of fI. Il/rida (Ill :lIld 17 Jllly 1990,5 July 1991), four C. zùz:alus (13-27 July 1991). two H. hcarld (4 and 5 August 1990), one H. affinis (24 July 1990). :md one A. sp/zagllicvla (I~ Jllly 1990) were also collected. In 1990 (15,17 July). two males of J-l. aequetilleta \\'cre collected in a canopy trap; these represent the first known nemctic collections of the males of this species. Baribeau and Maire (1983a) collected 105 tabanid males of 13 species in Malaise traps in bogs near Trois-Rivières, Québec. These included five of t'he species collected al Schefferville, including C. frigidtt.:>; H. fronralis, H. affinis. H. zonalis, and H. peellumani. In addition, Maire (1984b) collected a male C. frigidus in an emergence trap near Richmond Gulf. • Male H. arpadi, male H. zonalis, and males of a small dark Hybomitra sp. (hearlei?) were observed hovering approximately 20-30 cm above ground at an above-treeline hilltop location south of Schefferville on the morning of 17 July 1990. Two male H. zonalis had been collected the previous day l'rom the same location. 3. Seasonal Variation in Gonotrophic Age of Host-Seeking Females. A tot~' of 2,360 female tabanids of 10 species (8 Hybomitra, 2 Chrysops) were disseeted, and their parity and gonotrophic age determined. Based on • seasonal changes in parity and yolk deposition, populations of these species in the 110 Schefferville area appeared ta be either obligately anautagenous, facultatively • autagenous, or obligately autogenous. 3.1. Obligatcly Ànautogcnous Spccies. Typical of obligately anautogenous tabanid species, only nulliparous (N) female H. arpadi and H. aequetincta with Stage 1 and II ovarioles were co11eeted (Figs 6,8). Leprince and Maire (1990) reported a similar finding for these species in Richmond Gulf, and concluded that these species were anautogenous there. Thomas (1973a) aise reported anautogeny in H. arpadi populations in Alberta At the onset of the flight season, all female H. arpadi and H. aequetincta co11ected at Schefferville were nulliparous, but thereafter the proportion of individuals which were parous steadily increased (Figs. 6,8). This same trend was found by • McElligott and Ga110way (1991a) for, among other species, a population of host seeking H. arpadi in southeastern Manitoba. 3.1.1. Hybomitra aequetiru;ta. The seasonal emergence pattern of H. aequetincta at Schefferville, as inferred from the seasonal changes in the proportions of flies in different stages of gonotrophic developmem, fo11owed either of two distinct trends, depending upon site and year (Fig. 7). The pattern apparent in flies co11ected at Capricom was characterized by peak numbers of host-seeking females being co11ected during the first week of the • flight season (Fig. 7). During Week 1, a11 flies co11eeted were nulliparous (N), Il 1 over half of them \Vith ovarioles in Stage 1. Fernales \Vith NI ovarioles had • probably emerged recently and locally, since horse flies do no! begin yolk accumulation in oocytes until at least 24 hours post-ernergence (Magnarelli and Anderson 1977). The remainder of the flies collected during Week 1 had NIl ovarioles. NIl flies represent either individuals which had ernerged loca\ly, but more than 24 hours before their time of capture, or had flies which h:ld disperscd into the vicinity of the fen from elsewhere. By the second week of the flight season, ail H. aequetincta were still nulliparous, but the proportion of NI flies had declined considerably. This was indicative of reduced local emergence, since the decline in the proportion of NI flies was accompanied by an increase in the proportion of NIl flies. Collections of host-seeking flies during Week 2 were much lower than the • numbers collected during Week 1. This decline was probably attributable both to declining local emergence and 10 the dispersal of females away from their emergence sites in the fen, in search of blood meals. Vertebrate hosts (e.g., caribou Rangifer tarandus caribou (Gmelin); black bear Ursus americanus Pallas) were uncommon in the immediate vicinity of Capricom. Aiter Week 2, emergence, as indicated by the presence of NI flies, had ceased. By the third week of the flight season the first parous (P) individuals appeared arnong the NIl individuals which made up the host-seeking population. Parous flies represented those individuals who had either successfully obtained a blood meal eIsewhere but had oviposited locally, or had oviposited somewhere • eIse and bad dispersed into the vicinity of the fen in search of a second blood- 112 meal. The former case is more likely, owing to the predominance of recently • parous (e.g., parous sacculate (Psac) and parous yellow-bodied (PybI)) arnong the parous flies collected (Fig. 6). Psac ovarioles are ooly present in individuals less than 48 hours post-oviposition (Thomas 1972). Toward the end of the flight season (Weeks 3-6), nulliparous individuals gradually disappeared from the host seeking population. Presumably they either died off, gradually dispersed away from the fen, or successfully blood-fed, oviposited, and re-appeared as parous individuals. In any case, parous females gradually predominated late-season tabanid catches, as tabanid numbers declined toward the end of the flight sèason (Figs. 6,7). The above emergence pattern applied to H. aequetincta collected at Iron Arm in 1990 as well (Fig. 7). At Iron Arm in 1991, however, the seasonal • emergence trend was much different; catches of host-seeking female H. aequetincta did not peak until the third week of the flight season (Fig. 7). Numbers of NI females were highest in Week 2, although they were also present during weeks 1-3. This indicates that in 1991, emergence of H. aequetincta at Iron Arm was not maioly confined to the first week of the flight season, but rather was spread out over a three week period. Nil females predominated in the host seeking population during the first three weeks of the flight season. It is very possible that the predominance of NIl individuals during Weeks 1-3 represented a large influx of individuals who had emerged elsewhere (e.g., Capricorn) and dispersed into the vicinity of Iron Arm in search of hosts, rather than locally • emerged females colleeted within 24 hours of emergence. Il., After Week 3, parous individuals predominated (> 70%) among host • seeking H. aequetincca at a time when this species was still abundant (i.t:.. weekly average 155 flies trapped per day), rather than at a lime when the population was in decline at the end of the flight season (Figs. 6,7) This suggeslS that there was a large influ.x of blood-fed flies into the vicinity of Iron Arm from elsewhere, that these flies oviposited there, and then bt."gan seeking a second blood meal in the vicinity of the fen. The majority of parous individuals collected had Psac and PybI ovarioles, indicating recen!, probably local, oviposition. As at Capricorn, the majority of flies at the end of the flight season were parous (Fig. 7). There are numerous questions raised by the above results. Why, for example, was there apparently a great influx of flies in Iron Arm in 1990 and not • Capricom? Where did the influx of flies come from, how far do they travel, and where did they get their blood meals? These questions remain unanswered, and are likely ta remain 50 for the foreseeable future, owing to the tremendous logistic cost in designing research ta answer these questions, particularly in subarctic Québec-Labrador. A possible answer ta the question of why H. aequetincta emergence patterns varied at Iron Arm between 1990 and 1991 is suggested by one of the findings of the previous chapter on larval development. Immature C zinzaius and .C nigripes have a 3-4 year life cycle in the Schefferville area. It is very likely that the life cycle length of H. aequetincta and other subaretic Hybomitra species varies • as well. Take a case, for example, where the majority of late-instar larvae are at 114 slightly diffcrcnt stages of developmcnt when the substratc thaws in the spring, but • ail arc sufficiently large that they will pupate that spring. Pupation. and thus adult emcrgence, will be distributed over several weeks, and reflect the size distribution of the larvae coming out of hibernation. This would result in a seasonal abundance pattern similar to that observed at Iron Arm in 1990. If, however, the majority of mature larvae emerging from diapause in the spring are too small to pupate that year, they would continue developing until the following year. The following spring, ail of these larvae would be ready to pupate as soon as conditions were favourable. This, would lead to an emergence peak early in the f1ight season, as was observed at Iron Arm in 1991. The development of larval cohons presents a tantalizing explanation for a:mual and locally varying patterns of adult emergence of H. aequetincta at • Schefferville. During the present study, unfortunately, logistic limitations, coupled with low Hybomitra larva! abundance, prevented collection of data connecting seasona! larval size distribution with adult emergence patterns. A major problem in studying tabanid larvae is that they are very difficult to colleet, owing to their relatively diffuse distribution in a thick, fibrous substrate. Hopefully, in the future, a long-term study will be carried out in a locale where larval tabanids are sufficiently abundant that variation in their seasona! size distribution can be conneeted to the seasonal emergence patterns of the adults. • 115 3.1.2. Hybomitra arpadi • In ail but one C:lSe, trends in the se:lSonal changes in numbers of host seeking H. arpadi at different stages of ovarian development were similar ta either of the IWO patterns found for H. aequetincta (Figs. S,9). At Iron Arm and Caprico:11 in 1991, emergence peaked during the first week of the flight season, and numbers of host-seeking flies declined rapidly thereaftcr (Fig. 9). At Iron Arro in 199C, however, emergence W:lS spread out over severa) weeks, and numbers of host-seeking flies peaked in mid-se:lSon. In the 1990 flight se:lSon at Capricom, however, a combination of both early and mid-se:lSon emergence trends was apparent; numbers of NI females peaked in the first and fourth weeks flight season (Fig. 9). In light of the "larval size distribution" explanation of adult emergence patterns, this pattern would represent the combination of the IWO • emergence trends; one group of larvae would be sufficiently large to pupate as soon as condiùons permitted in the spring, whereas the other group would pupate over a period of weeks, as larval development permitted. Likely IWO cohort5 were involved. 3.2. Facultative!y Autogenous Species. Sorne female H. lurida and H. zonalis with NIII and NIV ovarioles were collected in addition to those individuals with NI and NIl ovarioles (Figs. 10,11). This indicated that at least sorne individuals of these species were facultatively autogenous in the Schefferville area, that is, they are capable of maturing eggs • without a blood meal, but will host-seek while the egg-batch is developing.- 116 3.2.1. Hybomi1ra lurida • Nulliparous females represented a significant proportion of host-seeking H. lurida only during the first " eek of the flight season (Fig. 10), and were only collected sporadically thereafter. It appears that H. lurida females in the Schefferville area mature their eggs very quickly, whether or not they have a blood meal. Many females had oviposited before host-seeking, and many were in the process of developing eggs while host-seeking, since even at the beginning of the flight season the majority of host-seeking females were parous or had NlII-IV ovarioles. A similar rapid increase in percent parity, and a high proportion of parous individuals present at the beginning of the flight season was repcrted by McElligott and Galloway (1991a) for H. lurida in Manitoba. They suspected facultative autogeny in this species. Thomas (1972,1973a), however, considered H. • lurida to be anautogenous in Alberta. At Richmond Gulf, Leprince and Maire (1990) found only one nulliparous individual among 14 H. lurida colleeted, but did not make any conclusions as to whether or not this species was anautogenou . because of the small sample size. 3.2.2. Hybomitra zonalis Unlike H. lurida, many nulliparous female H. zonalis were colleeted, particularly NI and NIl individuals. The number of host-seeking H. zonalis females generally peaked in Week 4 or 5 of the flight season, at the same time that numbers of NI females peaked (Figs. 11,12), suggesting that emergence was • spread out over the first four weeks, peaking in mid-season. Compared to H. lli arpadi and H. aequetincta. very few H. :onalis with NI ovarioles \Vere collected. • although this varied between years and sites (Figs. 11,12). Under certain conditions, a large proportion of female H. :onalis had apparently deve10ped NIl ovarioles by the time that they began their host-seeking activity. Apart from the presence of females \Vith NIII-IV follicles, the patterns of seasonal change in gonotrophic age of H. :onalis (Fig. 12) \Vere similar to that observed in both H. arpadi and H. aequetincta in years of mid·season abundance (Figs. 7,9). The presence of relatively fe\V NIII·IV femaies in the overall population indicates that sorne individuais of H. zonalis are facultatively autogenous, but the majority of females develop eggs anautogenously. At Iron Arro in 1991, nearIy aIl of the host-seeking flies collected had NIl ovarioles, and very few parous flies were collected (Fig. 12). This implies that either • survivorship of parous individuals is very low under sorne conditions, or that very few anautogenous individuals which had succeeded in finding a blood meal oviposited at the fen. At Richmond Gulf, Leprince and Maire (1990) found H. zonalis to be an anautogenous species, based upon their finding of nulliparous females. McElligotl and Galioway (1991a) also concluded that this was an anautogenous species, based upon changes in the proportion of parous individuais collected over the course of the flight season. • 118 3.3. Obligatcly Autogcnous Spccics. • Certain tabanid species in the Schefferville area \Vere apparently obligately autogenous for the first ovarian cycle, since only parous individuals \Vere collected; ail females of these species had therefore laid a first egg batch before they began host-seeking. Autogenous species included H. pechumani, H. hearlei, H. frontalis, H. astU/a, C. zinzalus, and C. nigripes. Thomas (1972), through lab rearing, found that H. frontalis \Vas autogenous in Alberta. Leprince and Maire (1990) concluded that H. hearlei and H. frontalis at Richmond Gulf \Vere autogenous based on a lack of nulliparous flies in field collections; similarly; 1 did not collect nulliparous females of H. frontalis at Churchill, Manitoba in 1989 (McElligott, unpubL data). This is the first report of autogeny in H. astuta, C. • zinzalus, and C. nigripes. 3.4. Other Species. Several species \Vere collected too seldom during this study for determination of seasonal changes in their ovarian development; these included C. aler, C. excitans, C. frigidus, C. furcatus, C. sordidus, H. affinis, H. Iiorhina, and A. sphagnicola. In Alberta, C. exdtans, C. jurcatus, and H. affinis are anautogenous, whereas Hybomitra Iiorlzina is autogenous (Thomas 1972,1973a). 1 found H. affinis to be anautogenous in southeastem Manitoba (McElligott and Galloway 1991a) • and northern Manitoba (McElligott, unpubL data). While C. arer is aulOgenous in Alberta (Thomas 1972. 1973a) and New • Hampshire (Lake and Burger 1980). this species is apP:lrently anautogenous in New York (Magnarelli 1976). Similarly. C. frigidus is alltogenolls in Albert:l and New Hampshire (Thomas 1972, 1973a. Lake and Burger llJi\O). but anmltogenolls in Ontario (Troubridge and Davies 1975) and New York (Magnarelli 197D). C/zrysops excitans is autogenous in Alberta (Thomas 1972,1973a), but anautogenous in New Hampshire (Lake and Burger 1980). It is likely that geographic variation exists in the ex-pression of aUlOgeny in these 3 $pecies. No data are available concerning aUlOgeny or :mautogeny in C. sordidus or A. sp/zagnicola. 4. Fat Body Deposition and Utilization at Successive Stages of Ovarian • Development. Patterns of fat body utilization and accumulation with successive stages of ovarian development were determined for seven tabanid species (Figs. 13-15). The proportion of females at each stage of ovarian development in each of the four fat body categories was remarkably consistent among species. For anautogenous species (H. arpadi, H. aequetincta) and facultatively autogenous species (H. zonalis, H. lurida), NI and NIl females consistently had more fat body than female with ovarioles in other stages of development (Figs. 13,14). NIII-IV females of facultatively autogenous H. zonalis and H. lurida, had less fat body than NIl females (Fig. 14), presumably because fat body was being channelled inta • their developing oôcytes. Unfortunately, no nulliparous individuals of obligately 120 autogenous speeies were eolleeted. When Bosler and Hansens (1974) reared • nuJ::parous females of obligately autogenous T. nigroviuarus from eggs, they found that the females possessed large amounts of fat body until they reaehed NIII, and that fat body was depleted as follicles matured. Parous flies had mueh less fat body, on average, than nulliparous individuals (Figs. 13-15). This \Vas presumably because fat body nutrients had been lost to egg production in the first ovarian cycle. Bosler and Hansens (1974) found that fat body \Vas low or exhausted in 90% of the parous female T. nigroviuarus they collected. There was a general trend of increasing fat body level with successive stages of ovarian development in nulliparous flies and, quite separately, in parous flies (Figs. 13,14). NIl females of H. arpadi, H. aequerincta and H. zonalis • appeared to have slightly more fat, on average, than those with NI ovarioles. Similary, Psac females of H. arpadi, H. aequerincta, H. zonalis. H. lurida and H. pechumani had slightly more less fat body, on average, than PybI females, who in turn had less fat body that PybIl fema!es. :It is possible that the apparent increase in the level of fat body bel\veen Stages 1 and II of both parous and nulliparous flies was a consequence of deposition of nutrients obtained from carbohydrate (i.e., nectar) feeding. This contradicts the findings of Auroi and Briegel (1985) that females of the horse fly Haematopota pluvialis (Linneus) does not synthesize net energetic reserve material in substantial amounts, either as fat or as glycogen, and that females require frequent feeding on carbohydrate sources is necessary to • sustain life. Magnarelli (1985), however, noted that caloric reserves of C 1::1 fuliginoslIs did not decline conistantly o\'er the duration of the flight season. as • Auroi and Briegel (1985) found with H. pILH'ialis. but rather tluctuated. Mosquitoes are known to synthesize fat body from carbohydrate meals (V:m Handel 1965), and il is possible that sorne t:\h:mid species do so as weil. • • 122 LITERATURE CITED. • Alverson, D.R., and R. Noblet. 1977. Activity of female Tabanidae (Diptera) in relation to selected meteorol09ical factors in South Carolina. J. Med. Entomol. 14: 197-200. Anderson, J.F. 1971. Aut0geny and mating and their relationship to biting in the saltmarsh deer fly, Chrysops atlanticus (Diptera: Tabanidae). Ann. Entomol. Soc. Amer. 64: 1421·1424. Anderson, J.R., W. Olkowski, and J.B. Hoy. 1974. The response of tabanid species to COz baited insect flight traps in northern California. Pan. Pac. Entomol. 50: 255-268. Auroi, C. 1982. Physiological age of Tabanid (Diptera) populations in Switzerland. J. Med. Entomol. 19: 281-284. Auroi, C., and H. Briegel. 1985. Energy reserves and carbohydrate feeding in field caught, female horse flies (Haematopota pluvia7is) (Diptera: Tabanidae). Myia 3: 403-414. Bailey, N.S. 1948. The hovering and mating behavior of Tabanidae: a review of the literature with some original observations. Ann. Entom07. Soc. Amer. 41: 403-412. Baribeau, L., and A. Maire. 1983a. Spatial distribution of Tabanidae (adults and larvae) in two bogs in southern Quebec. Mosq. News 43: 24-29. • Baribeau, L., and A. Maire. 1983b. Abundance and seasonal distribution of Tabanidae in a temperate and in a subarctic locality of Quebec. Mosq. News 43: 135-143. Beesley, W.N., and W. Crewe. 1963. The bionomics of Chrysops si7acea Austen. II. The biting rhythm and dispersal in rain forests. Ann. Trop. Med. Parasit07. 57: 191-203. Bennett, G.F., and S.M. Smith. 1968 .. Phosphorus 3Z for marking Tabanidae (Diptera). Mosq. News 28: 559-569. Blickle, R.L. 1955. Feeding habits of Tabanidae. Entom07. News 66: 77-78. Blickle, R.L. 1959. Observations on the hovering and mating of Tabanus bishoppi Stone (Diptera: Tabanidae). Ann. Entom07. Soc. Amer. 52: 183-190. Bosler, E.M., and E.J. Hansens. 1974. Natural feeding behavior of adult saltmarsh greenheads and its relation to oogenesis. Ann. Entom07. Soc. Amer. 67: 321-324. Burgess, N.R.H., A.E. Shuttleworth, and K.N. Chetwyn. 1979. Adhesive panels as a control method for the common cleg, Haematopota p7uvia7is L. • (Diptera: Tabanidae). J. Roya7. Army Med. Corps 125: 156-161. Burnett, A.M., and K.L. Hays. 1974. Some influences of meteorological factors on the flight activity of female horse flies (Diptera: Tabanidae). • Environ. Entomo7. 3: 515-521. Catts, E.P. 1970. A canopy trap for collecting Tabanidae. Mosq. News 30: 472-474 Catts, E.P., and W. Olkowski. 1972. Biology of Tabanidae (Diptera): matin9 and feeding behavior of Chrysops fu7iginosus. Environ. Entomo7. 1: 448-453. Christophers, S.L. 1911. The development of the egg follicle in anophelines. Pa7udism 2: 73-89. Chvala, M., L. Lyneborg, and J. Moucha. 1972. The Horse F7ies of Europe. Entomol. Soc. Copenhagen publ. Ch. 7. Clark, G.G., C.P. Hibler, B.R. Donaldson, and G.H. Gates. 1976. Haematophagous activities of Hybomitra 7aticornis and H. tetrica rubriIata (Diptera: Tabanidae). J. Med. Entomo7. 13: 375-377. Cooksey, L.M., and R.E. Wright. 1987. Flight range and dispersal ability of the host-seeking horse fly, Tabanus abactor (Diptera: Tabanidae) in north-central Oklahoma. Environ. Entomo7. 16:211-217. Corbet, P.S., and A.J. Haddow. 1962. Diptera swarming high above the forest canopy in Uganda, with special reference to Tabanidae. Trans. Roya 7 Entomol. Soc. London 114: 267-284. • Dale, W.E., and R.C. Axtell. 1975. Flight of the salt marsh Tabanidae (Diptera), Tabanus nigrovittatus, Chrysops atIanticus, and C. fu7iginosus: correlation with temperature, light, moisture, and wind velocity. J. Med. Entomo7. 12: 551-557. Dale, W.E., and R.C. Axtell. 1976. Salt marsh Tabanidae (Diptera): a comparison of the abundance and distribution in 5partina and Juncus habitats. J. Med. Entomo7. 12: 671-678. Davies, D.M. 1959. Seasonal variation in tabanids (Diptera) in Algonquin Park; Ontario. Cano Entomo7. 41: 548-553. Davey, J.T., and F.J. O'Rourke. 1951. Observations on Chrysops si7acea and C. dimidiata at Benin, soutnern Nigeria. Ann. Trop. Med. Parasito7. 45: 66. Detinova, T.S. 1962. Age grauping methads in Diptera af medica7 impartance. World Health Organ., Geneva. Dukes, J.C., T.D. Edwards, and R.C. Axtell. 1974. Distribution of larval Tabanidae in a Spartina a7ternif7ara salt marsh. J. Med. Entama7. Il: 79-83. • Foil, L. 1989. Tabanids as vectors of disease agents. Parasita7. Taday 5: 88·95. Foil, L., W.V. Adams Jr., C.J. Issel, and R. Pierce. 1984. Tabanid (Diptera) populations associated with an Equine Infectiuus Anemia outbreak in an • apparently infected herd of horses. J. Med. Entomol. 21: 28-30. Freeman, T.N. 1953. Interim report of the distribution of the tabanids obtained in the Northern Insect Survey. Def. Res. Bd. Cano Environ. Proto Tech. Rep. 4. 3 pp. + 35 maps. Golini, V.l., and R.E. Wright. 1978. Relative abundance and seasonal distribution of Tabanidae (Diptera) near Guelph, Ontario. Cano Entomol. 110: 385-398. Gordon, R.M., L.J. Chwatt, and C.M. Jones. 1948. The results of a preliminary entomological survey of loiasis at Kumba, British Cameroons, together with a description of the breeding places of the vectro and suggestions for future research and possible methods of control. Ann. Trop. Med. Parasitol. 42: 364-376. Hanec, W., and G.K. Bracken. 1962. Response of horse flies (Diptera: .Tabanidae) to light. Ann. Entomol. Soc. Amer. 55: 720-721. Hanec, W., and G.K. Bracken. 1964. Seasonal and geographic distribution of Tabanidae (Diptera) in Manitoba, based upon females captured in traps. Cano Entomol. 96: 1362-1639. Hocking, B. 1952. Protection from northern biting flies. Mosq. News. 12: 91-102. • Hocking, B. 1953. On the intrinsic flight range and speed of flight of insects. Trans. Entomol. Soc. London 104: 1-21. Hollander, A.L., and R.E. Wright. 1980. Daily activity cycles of eight species of Oklahoma Tabanidae (Diptera). Environ. Entomol. 9: 600-604. James, H.J. 1953. Natural control of Tabanidae (Diptera) in the region of Churchill, Manitoba. Ann. Rep. Entomol. Soc. Ont. 82: 70-74. Jamnback, H. 1969. Bloodsucking flies and other nuisance arthropods of New York State. N.Y. State Mus. and Sei. Serv. Mem. 19: 1-90. Joyce, J.M. Jr., and E.J. Hansens. 1968. The influence of weather on the activity and behavior of greenhead flies, Tabanus nigrovittatus Macquart and Tabanus lineola Fabricius. J. N.Y. Entomo7. Soc. 76: 72-80. Judd, W.W. Lj67. lnsects From McConnell River, N.W.T. Entomo7. News 78:50-55. Kingston, S.R., J.K. Wangberg, and D.P. Sanders. 1986. Flight behavior and nocturnal resting sites of Tabanus abactor Philip (Diptera:Tabanidae) in the Texas Rolling Plains. J. Kansas. Entomo7. Soc. 52: 338-342. Kniepert, F.W. 1982. Der einfluss verschiedener witterungsfactoren auf die flugaktivitat der bremsen (Diptera: Tabanidae). Z. Ang. Entomol. 93: • 191-207. Krinsky, W.L. 1976. Animal disease agents transmitted by horse flies and deel' • flies (Diptera:Tabanidae). J. Hed. Entomol. 13: 225-275. Lake, D.J., and J.F. Burger. 1980. Ovarian development in adult Chrysops (Diptera: Tabanidae) in northern New England, with emphasis on Chrysops ater and C. mitis. J. Hed. Entomol· 17: 502-505. Lancaster, J.L., and M.V. Meisch. 1986. Arthropods in livestock and poultry production. Ellis Harwood Ltd., West Sussex, Engl. pp. 19-38. Lane, R.S., and J.R. Anderson. 1982. The reproductive life history and blood me al sources of Chrysops hirsuticallus (Diptera: Tabanidae). J. Med. Entomol. 19: 157-163. Lane, R.S., and J.R. Anderson. 1983. Multiple autogeny and burrow oviposition by a marine horse fly (Diptera: Tabanidae). J. Med. Entomol. 20: 212-213. Lanp, R.S., J.R. Anderson, and C.B. Philip. 1983. Biology of autogenous horse flies native to coastal California: Apatolestes actites (Diptera: Tabanidae). Ann. Entomol. Soc. Amer. 76: 559-571. Leprince, D.J. 1989. Gonotrophic status, sperm presence, and sugar feeding patterns in southwestern Quebec tabanid (Diptera) populations. J. Amer. Mosq. Cont. Ass. 5: 383-386 . Leprince, D.J., and M. Bigras-Poulin. 1990. Gonotrophic status, sperm presence, and sugar feeding patterns in a Hybomitra lasiophthalma population • (Diptera: Tabanidae). J. Med. Entomol. 27: 31-35. Leprince, D.J., and D.J. Lewis. 1983. Aspects of the biology of female Chrysops univittatus (Diptera:Tabanidae) ;n southwestern Quebec. Cano Entomol. 115: 421-425. Leprince, D.J., and D.J. Lewis. 1986. Sperm presence and sugar feeding patterns in nulliparous and parous Tabanus quinquevittatus Weidmann in southwestern Quebec. Ann. Entomol. Soc. Amer. 79: 912-917. Leprince, D.J., D.J. Lewis, and J. Parent. 1983. Biology of male tabanids (Diptera) aggregated on a muntain summit in southwestern Quebec. J. Med. Entomol. 20: 608-613. Leprince, D.J., and A. Maire. 1990. Parity, stage of follicular development, and sperm presence in hemiarctic host-seeking Hybomitra populations (Diptera: Tabanidae). J. Med. Entomol. 27: 835-838. Lewis, D.J. 1960. Notes on Chrysops bicolour (Diptera: Tabanidae) in Tanganyika. Proc. Roy. Soc. London (A) 35:75-78. Lewis, D.J. 1987. Biting flies (Diptera) of peatlands and marshes in Canada. pp. 133-140. In Rosenberg, D.M. and H.V. Danks. Aquatic insects of • peatlands and marshes in Canada. Mem. Entomol. Soc. Cano 140: 174 pp . 126 Lewis, D.J., and G.;'. Bennett. 1977. Bitin9 flies of the eastern Maritime provinces of Canada. 1. Tabanidae. Cano J. Zool. 55: 1493-1503. • ~orse Lewis, D.J., and D.J. Leprince. 1981. flies and deer flies (Diptera: Tabanidae) feeding on cattle in southwestern Quebec. Cano Entomol. 113: 883-886. Logothetis, C., and H.H. Schwardt. 1948. Biological studies on the horse flies of New York. J. Econ. Entomol. 41: 335-336. Magnarelli, L.A. 1976. Physiological age of Tabanidae (Diptera) in Eastern New York State, U.S.A. J. Med. Entomol. 12: 679-682. Magnarelli, L.A. 1985a. Hovering and swarming behavior of males of Hybomitra lasiophthalma and Hybomitra aurilimba (Diptera: Tabanidae). Myia 3: 475-484. Magnarelli, L.A. 1985b. Blood feeding and oviposition by tabanids (Diptera) in the laboratory. J. Med. Entomol. 22: 600-603. Magnarelli, L.A. 1985c. Calorie reserves in salt marsh deer flies Chrysops fuliginosus (Diptera: Tabanidae). Ann. Entomol. Soc. Amer. 78: 594-599. Magnarelli, L.A. 1987. Parity and sugar feeding in Atylotus bicoiour (Diptera: Tabanidae). J. N.Y. Entomol. Soc. 95: 487-490. Magnarelli, L.A., and J.F. Anderson. 1977. Follicular development in salt marsh Tabanidae (Diptera) and incidence of nectar feeding in relation to • gonotrophic activity. Ann. Entomol. Soc. Amer. 70: 529-533. Magnarelli, L.A., J.F. Anderson, and J. H. Thorne. 1979. Diurnal nectar feeding of salt marsh Tabanidae (Diptera). Environ. Entomol. 8: 544-548. Magnarelli, L.A., and J.F. Anderson. 1980. Feeding behavior of Tabanidae (Diptera) on cattle and serologie analysis of partial blood meals. Environ. Entomol. 9: 664-667. Magnarelli, L.A., and J.F. Anderson. 1981. Sugar feeding by female tabanids (Diptera: Tabanidae) and its relation to go~otrophic activity. J. Med. Entomol. 18: 429-433. Magnarelli, L.A., and J.G. Stoffolano Jr. 1980. Blood feeding, oogenesis, and oviposition by Tabanus nigrovittatus in the laboratory. Ann. Entomol. Soc. Amer. 73: 14-17. Magnarelli, L.A., and L.L. Pechuman. 1975. Ovarian studies of Tabanus quinquevittatus (Diptera: Tabanidae). J. Med. Entomol. Il: 687-690. Maire, A. 1984a. Les Tabanidae (Diptera) à la limite septentrionale des forêts (domaine maritime du Nouveau-Québec). Cano Entomol. 116: 227-233. Maire, A. 1984b. Les taons (Diptera: Tabanidae) d'une tourbière à palses située • à la limite des forêts (Nouveau-Québec). Cano Entomol. 116: 1679-1682. '~ l -' Maire, A., and J-P. Beaudoin. 1984. Hybamitra aeqüetincta et H. zana7is e· (Diptera: Tabanidae): clé d'identification, distribution et abondonce saisonnière au Québec-Labrador. Rev. Entama7. Qué. 29: 62-72. Martin, J.E. 1977. Collecting, preparing, and preserving insects. mites. and spiders. The insects and arachnids af Canada. Part 1. Ministry of Suppl Y and Services, Canada. Ottawa, Ontario. 182 pp. McElligott, P.E., and T.D. Galloway. 1991. Seasonal distribution and parity of host-seeking horse flies (Diptera: Tabanidae) from a pasture near Seven Sisters, Manitoba. Cano Entama7. 123: 361-370. McElligott, P.E., and T.D. Galloway. 1991. Daily activity of horse flies (Diptera: Tabanidae: Hybamitra spp.) in northern and southern Manitoba. Cano Entama7. 123: 371-378. McElligott, P.E., and S.B. Mc1ver. 1987. Range of attractiveness of carbon dioxide to Hybamitra spp. (Diptera: Tabanidae). J. Amer. Masq. Canto Ass. 3: 655-656. Mer, G.G. 1936. Experimental study on the development of the ovary in Anaphe7es e7utus Edw. (Dipt. Culic.). Bu77. EntamaI. Res. 27: 351-359. Miller, L.A. 1951. Observations on the bionomics of sorne northern species of Tabanidae (Diptera). Cano J. Zaal. 29: 240-263 . Morris, C.D., and G.R. Defoliart. 1971. Seasonal parous rates in Hybamitra • 7asiaphthalma (Diptera: Tabanidae). J. Med. Entama7. 8: 207-208. O'Meara, G.F., and G.B. Craig, Jr. 1969. Monofactorial inheritance of autogeny in Aedes atrapa7pus. Masq. News 29: 14-22. Pechuman, L.L. 1981. The horse flies and deer flies of New York (Diptera: Tabanidae). Second edition. Search Agricu7ture 18: 1-71. Pechuman, L.L., and J.J.S. Burton. 1969. Seasonal distribution of Tabanidae (Diptera) at Texas Hollow, New York in 1968. Masq. News 29: 216-220. Pechuman, L.L., H.J. Teskey, and D.M. Davies. 1961. The Tabanidae (Diptera) of Ontario. Prac. Entama7. Sac. Ont. 91: 77-121. Pechuman, L.L., D.W. Webb, and H.J. Teskey. 1983. The Diptera, or true flies of Illinois. 1. Tabanidae. ~17inais Nat. Hist. Surv. Bu77. 33: 1-121. Perich, M,J., R.E. Wright, and K.S. Lusby. 1986. Impact of horse flies (Diptera: Tabanidae) on beef cattle. J. Ecan. EntamaI. 79: 128-131. Philip, C.B. 1978. An astonishing hovering record of males of the western horse fly Pilimas califarnica (Bigot). Pan Pac. EntamaI. 54: 74. Raybould, J.N. 1967. Studies on Chrysops bica70r Cordier (Diptera: Tabanidae) at Amani, Tanzania, with particular reference to feeding behavior. Ann. • Trap. Med. ParasitaI. 61: 167-114. 12S Roberts, R.H. 1966. Biological studies of Tabanidae. 1. Induced oviposition. • Mosq. News 26: 435-438. Roberts, R.H. 1969. Biological studies of Tabanidae Il. A preliminary study of female tabanids attracted to a bait animal. Mosq. News 29: 236-238. Roberts, R.H. 1971. Effect of amount of COz on collection of Tabanidae in Malaise traps. Mosq. News 31: 551-558. Roberts, R.H. 1975. Relationship between amount of COz and the collection of Tabanidae in Malaise traps. Masq. News 35: 150-155. Roberts, R.H. 1976. The comparative efficiency of six trap types for the collection of Tabanidae (Diptera). Masq. News 36: 530-534. Rockel, E.G. 1969. Autogeny in the deer fly Chrysaps fuliginasus (Diptera: Tabanidae). J. Med. EntamaI. 6: 140-142. Rockel, E.G., and E.J. Hansens. 1970. Emergence and flight activity of salt marsh horse flies and deer flies. Ann. Entamol. Sac. Amer. 63: 27-31. Sheppard, C., and S.H. Wilson. 1976. Flight range of Tabanidae in a Louisiana bottomland forest. Environ. EntamaI. 5: 752-754. Sheppard, C., and S.H. Wilson. 1977. Relationship of horse fly host-seeking activity to the edge of wooded areas in southern Louisiana. Enviran. EntamaI. 6: 781-782. • Sheppard, C., S.H. Wilson, and S.R. Farthing. 1980. Flight routes of Tabanidae ___ in a Louisiana bottomland hardwood forest. Enviran. Entomal. 9: 489-491. Smith, S.M. 1970. The biting flies of the Saker Lake region, Northwest Terl'itories (Diptera: Culicidae and Simuliidae). Unpubl. Ph.D. Thesis, University of Manitoba, Winnipeg. 280 pp. Smith, S.M., D.M. Davies, and V.I. Golini. 1970. A contribution to the bionomics of the Tabanidae (Diptera) of Algonquin Park, Ontario: seasonal distribution, habitat preferences, and biting records. Cano EntamaI. 102: 1461-1473. Spielman, A. 1957. The inheritance of autogeny in the Culex pipiens complex of mosquitoes. Am. J. Hyg. 65: 404-425. Spielman, A. 1957. Sionomics of- autogenous mosquitoes. Ann. Rev. Entamel. 16: 231-248. Surcouf, J.H. 1908. Les tabanides de France et des pays Limitrophes. Encycl. EntamaI. Sér. A5. Lechevalier, Paris. 261 pp. Tashiro, H., and H.H. Schwardt. 1949. The biology of the major species of • horse flies in central New York. J. Ecan. EntamaI. 42: 269-272. Teskey, H.J. 1969. Larvae and pupae of sorne eastern North American Tabanidac • (Diptera). Hem. Entomoi. Soc. Cano 63: 1-147. Teskey, H.J. 1991. The horse flies and deer flies of Canada and Alaska (Diptera: Tabanidae). The insects and arachnids of Canada. Part 16. Ministry of Suppl y and Services, Canada. Ottawa, Ontario. 381 pp. Thibault, J., and P.P. Harper. 1983. Les peuplements de taons (Diptères: Tabanidae) d'une forêt des basses-Laurentides: inventaire, phénologie, activité et habitats. Nat. Cano 110: 27-36. Thomas, A.W. 1971. Autogeny and anautogeny in sorne species of tabanids (Diptera: Tabanidae) in Alberta, Canada. Unpubl. Ph.D. Thesis, University of Alberta, Edmonton. 108 pp. Thomas, A.W. 1972. Physiological age structure of adult tabanid populations (Diptera: Tabanidae) in Alberta, Canada. J. Hed. Entomo7. 9: 295-300. Thomas, A.W. 1973a. Follicle developmental stages in blood-sucking horse flies (Diptera: Tabanidae) in Alberta, Canada. J. Hed. Entomo7. 10: 325-328. Thomas, A.W. 1973b. The deer flies (Diptera: Tabanidae: Chrysops) of Alberta. Quaest. Entome7. 9: 161-171. Thompson, P.H., J. Holmes, and T.I. Araujo. 1979a. Determination of parity in populations of dominant Tabanus species (Diptera: Tabanidae) in southeast Texas. Southw~,L. Entomo7. 4: 181-191. • Thompson, P.H., J. Holmes, and T.I. Araujo. 1979b. Parity in Texas populations Hybomitra 7asiophtha7ma (Macquart). Southwest. Entomol. 4: 273-277. Thompson, P.H., B.F. Hogan, and H. Oel Var Petersen. 1980. Rearing of Texas Tabanidae (Diptera). III. Tra~~ing, survivorship, and limited rearing of Hybomitra 7asiophtha7ma (Macquart). Southwest. Entomo7. 5: 191-195. Thornhill, A.R., J.P. Gibert, and K.L. Hays. 1971. How far do horse flies and deer flies fly? High7ights of Agric. Res. 18: 1. Thornhill, A.R., and K.L. Hays. 1972. Dispersal and flight activities of sorne species of Tabanus (Diptera: Tabanidae). Environ. Entomo7. 1: 602-606. Townes, H. 1972. A light-weight Malaise trap. Entomol. News. 83: 239-247. Troubridge, D.A., and D.M. Davies. 1975. Seasonal changes in physiological age composition of tabanid (Diptera) populations in southwestern Ontario. J. Med. Entomo7. 12: 453-457. Twinn, C.R., B. Hocking, W.C. McDuffie, and H.F. Cross. 1948. A preliminary account of the biting flies at Churchill, Manitoba. Cano J. Res. (Sec. Dl 26: 334-357. • Van Handel, E. 1965. The obese mosquito. J.Physio7. 181: 478-486. 130 Weber, N.A. 1950. A survey of insects and related arthropods of arctic • Alaska. Trans. Amer. EntamaI. Soc. 76: 147-206. White, D.J., C.D. Morris, and K. Green. 1985. Seasonal distribution of New York anthropophilic Tabanidae (Diptera) and observations on the dispersal of several species. Environ. EntamaI. 14: 187-192. Wilkerson, R.C., J.F. Butler, and L.L. Pechuman. 1985. Swarming, hovering and mating behavior of male horse flies and deer flies (Diptera: Tabanidae). Myia 3: 515-546. Wilson, B.H. 1967. Feeding, mating, and oviposition studies of the horse flies Tabanus lineola and T. fuscicostatus (Diptera:Tabanidae). Ann. EntamaI. Sac. Amer. 60: 1102-1106. • • • • • Tobie 1. Spcclcs, relative obundanccs, and dates of obscrvcd fllght seasons of tabanfd spccfes collcctcd in 1990 and 1991 'rom Iron Arm and Caprlcorn Icns, ncDr Schefferville, Ouchec. 1990 1991 Iron Am Coprfcorn Iron Am Cnpricorn Spccles n Dotes n Dotes n Dates n Dotes -- labanlMc: H'r'boml tra Bequet ineta 1862 3 Jul 6 Aug 921 2 Jul 6Aug 2967 4 Jul 7 Aug 3484 4 Jul 11 Au!) H. nff lois 8 3 Jul 17 Jul 2 4 Jul • 25 Jul 9 14 Jul 4 Aug 6 6 Jul 30 Jul H. nrpadi 7983 3 Jul • 12 Aug 2483 2 Jul 15 Aug 3799 4 Jul 11 Aug 1120 4 Jul 10 Aug H. astuta 32 17 Jul 6 Aug 26 23 Jul 15 Aug 13 27 Jul 11 Aug 24 26 Jul 11 Aug H. frontails 45 16 Jul 6 Aug 95 16 Jul 15 Aug 13 25 Jul 10 Aug 66 26 Jul 9 Aug H. hearlei 267 9 Jul 3 Aug 176 14 Jul • 6Aug 26 13 Jul 4 /lug 110 14 Jul 6 Aug H. 1iorhina 2 1 Aug 5 Aug 4 1 Aug • 6Aug 0 6 31 Jul 11 Aug H. lurida 124 3 Jul 24 Jul 80 25 Jun • 26 Jul 296 4 Jul 4 Aug 60 4 Jul 6 Aug H. pçchU'l\llni 75 22 Jul 16 Aug 42 24 Jul 13 Aug 30 27 Jul 11 Aug 40 30 Jul 11 Aug H. zonal is 779 3 Jul 8 Aug 612 2 Jul 15 Aug 359 4 Jul 10 Aug 430 5 Jul 11 Aug Atrlotus sph8qnicolus 5 17 Jul 24 Jul 8 23 Jul 26 Jul 4 21 Jul 4 Aug 1 29 Jut Total: 11182 4449 7516 5347 Chrysoplnae: Chrvsops ater 55 16 Jul 6 Aug 28 15 Jul 3 Aug 19 7 Jul 4 Aug 10 26 Jul 6 Aug C. excitans 9 3 Jul 24 Jul 5 16 Jul 25 Jul 3 6 Jul • 7 Jul 2 5 Ju' 20 Jul C. furcatus 72 16 Jul 6 ,~'~g 2 24 Jul 6Aug 16 25 Jul 5 Aug 0 C. nigripes 65 9 Jul 25 ,ul 42 14 Jul 28 Jul 21 16 Jul 5 Aug 81 5 Jul 6 Aug C. sordldus 3 18 Jul 19 Jul 10 24 Jul 25 Jul 2 15 Jul 25 Jul 0 C. tlnzalus 226 15 Jul • 6 Aug 109 15 Jul • 3Aug 237 12 Jul 10 Aug 184 25 Jul II Aug Total: 430 196 298 277 Oversll Total: 11612 3 Jul • 16 Aug 4763 25 Jun • 15 Aug 7814 4 Jul • 11 Aug 5624 4 Jul • 11 Aug • Figure 1. Map of vicinity of Schefferville, Quebec, indicating • locations of Iron Arm fen and Cap~icorn fen. • • • ~. . ~ .Adr0.1 Lal~.}S; " \;.,. '.' 1 ~ "'u,"~er: ;ncJicc'e 'oeclio" oF sampte- o ~ ID I~ ~...... 1 , : 1 b:Io:';·;~:1 to'es . 1(;,10"''''''' t6'}î S"tcmp ona ,,;)n-w~odC'~mud:e-g lmm~ul Lond obor Fi9ure 2. Seasonal abundance of Hybomitra spp. at Iron Arm fen in 1990 • and 1991. • • ~ ~ ~ s..,. 0 - (fi dl "0 ;;: dl <,' 1) -:> 0 ~ Cl - :> ci ë • o m m. ~ • • 134 Figure 3. Seasonal abundance of Hybomitra spp. at Capri corn fen in 1990 • and 1991. • i"'j"'I"'j"'I'ill"II"'i'''I'''I,lllil'I''IJ CD "0 '" CD ii= '0 -0 .!li <;; ci ë5 :J c: () Cl «:J Cl 1 -Cl >. "5-, .l!! '~ ~ ~ '0 'c:: ~ '- CD c:: ~ § ~ ::. ~ CD .t:: • t:T 2 .t:: () CD CD III Cl. <;; :J 01 «:J o Cl -Cl • • 135 Figure 4. Seasonal abundance of Chrysops spp. at Iron Arm fen in 1990 • and 1991. • ~ l< li: ~ 0 ... ~ • III a Cl) ~-... • • 136 Figure 5. Se as on al abundance of Chrysops spp. at Capri corn fen in 1990 • and 1991. • • • • • , - .~ "0 '" CD Li: Ü • ~ '0 "0 ci U Z <0 -:J 01 :J • (j) :J Cl :J o Cl ...Cl • • 13ï Figure 6. Proportions of host-seeking female Hybomitra aequetincta collected dt weeks 1-6 of the flight season with ovarioles at successi\'e staof>~ of development (Numbers above bars indicate • mean numbers of females collected per week). • 37 ,,, .231 1S• • r1l'"bl! r-jl'vlil ["11':-.11.: ""1 LlNI! • .NI '" 60 Iron Arm Fen 1990 ...• , '" :10 .. - .. ~.~.,' .' .' .~ ... Q • 41 •• .. .. ., . - .'. '; Caprlcern Fon 1990 "0 ..... 0=.. ..~.. ..E ... 107 .. •• u. '00 • ëi ë.. !!.. '" ~ "" Iron Arm Fen 1991 '" eGO lO' •• •• • Caprlcern Fon 1991 • • Figure 7. Average numbers of host-seeking female Hybomitra aequetincta collected during weeks 1-6 of its flight season with Nulliparous Stage l, Nulliparous Stage II, and Parous • ovarioles • n Parou:-. 250 o NlIliip.UOIiS Il • o Nullip'lIotl~ 1 200 150 Iron Arrn, 1990 100 300 250 Capricom, 1990 ~ i'" ëil'" E u.'" 500 • zci 1 Iron Arrn. 1991 700 600 500 Capricom, 1991 400 300 100 0 1 2 3 4 5 6 • Week of Flight Season • 139 Figure 8. Proportions of host-seeking female Hybomitra arpadi collected at weeks 1-6 of the flight season with ovarioles at successive stages of development (Numbers above bars • indicate mean numbers of females collected per week). • •• 1010 ." 100 '" • l~li'ybll UI'ybl oc- []P:-.IC • Nil o .NI 17. .7 '00 • .. .- C~IoOfn F.n) QQO .., c... ~ '".. .. ,~ ,. ..E .. • .... '00 ë ë.. 1:! .' '1' c.... ':~"",)o.-~~ Iton Arm Fen 1G~l .. :n •• • '00.. 60 C.pr'oom Fen 1GG1 ~~~:~!: ~~fé;' • • w..... __ • 140 Figure 9. Average numbers of host-seeking female Hybomitra arpadi collected during weeks 1-6 of its flight season with Nulliparous Stage l, Nulliparous Stage II, and Parous • ovarioles • [] l\uoLl:-' 1.200 [J Nullip.Il11Il:'- 11 U Nlllllp'UllLl:'\ 1 • 1.000 . 800' 600 . Iron Arm. 1990 a 250 200 150 Capricom. 1990 100 50 -'<: Q) Q) ..' , . ~ a Q) ï1l E 500 u.Q) ci 400 • Z dl 300 > Iron Arm. 1991 100 ,' ... ~.... 0 100 Capricorn, 1991 oLS1.:.28i::=:=====--.L__---'-__~ 1 2 3 4 5 6 • Week of Flighl Season • 141 Figure 10. Proportions of host-seeking female Hybomitra 7urida collected at weeks 1-6 of the flight season with ovarioles at successive stages of development (Numbers above bars • indicate mean numbers of females collected per week). • 11 OPybli 100 • • DPybl ClP:;'.lc DNIII • 00 o Nil .NI Iron Arm Fen 1990 '" :.<:::: ~,[Jl~ii~i:i~~:i:~~i~i:Lt::ti::~:i!~:ii~:ii: J=;::il'.~"i' .~.~. L_.....__-->__ :zo 2 12 • '00 00 60 Caprlcern Fan 1990 ::::::::::::: '" -0 Q "" Q, Q, ,::'" 0 .5!.. '"E ff1 Q • .... '00 • '0 'E Q 00 ~ Q C- OC Iron Arm Fon 1991 '" "" 0 ,. '00 00 OC Caprlcern Fon 1991 '" "" I~~~;~~;; f~?%&'f 0 2 3 • 5 • • w.....__ '~ • l ~- Figure 11. Proportions of host-seeking female Hybomitra zona7is collected at weeks 1-6 of the flight season with ovarioles at successive stages of development (Numbers above bars • indicate mean numbers of females collected per week) .. • , ,. •• 1;:7 • LlI'ybll 'CO l'I'ybl LlP:<-.IC UNIII LI Nil • '" .NI 60 <0 Iron Arm Fen' 990 20 ~'" 0 • 143 Figure 12. Average numbers of host-seeking female Hybomitra arpadi collected during weeks 1-6 of its flight season with Nulliparous Stage l, Nulliparous Stage II, and Parous • ovarioles collected per week. • [l P:UOU:io 140 [] Nllllip'lIou::.111 Nul1i~l'll OU:io Il 120 o o Nlillip.u IJU:io 1 • 100 80 80 Iron Arm, '990 40 20 0 140 120 100 80 80 Capricom, '990 ~ '" 40 20 ~'" '.::~~':.:''-:';.. CI) .''::-'. ::'\.'.. #,....:. Cil'" 0 E "-'" ci 140 Z dl 120 • > <: 100 80 Iron Arm, 1991 80 40 20 0 50 40 30 Capricom, 1991 20 10 0 1 2 3 4 5 6 • Week of Flight Season • l ~~', Figure 13. Proportions of host-seeking females of obligately anautogenous species (i.e., Hybomitra aequetincta and H. arpadi) at successive stages of ovarian development with • abdomens containing four different quantities of fat body. • • Hybomitra atpI/di lCO Do 01 .. llQ 1:]2 ..co .3 E ...co '" '0 ë co ca ~ co Co. "" 0 NI Nil NIII Psac Pybl Pybll Follicular Stage Hybomitm s8quetincls • 'co .. llCl ..co E ...co '" '0 ë ca ~ co Co. "" 0 NI NIl NIII Psac Pybl Pybll Follicular Stage • • Figure 14. Proportions of host-seeking females of facultatively autogenous species (i.e., Hybomitra 7urida and H. zona7is) at successive stages of ovarian development with abdomens • containing four different quantities of fat body. • • 1rlbomitra /urida .'" 00 .. ~.:- ' , . D1 ...... :...::. D2 llQ ~~1r:i;.: ~ .3 " ':.~.. '-- . .. ~§~.?'. E 60 ..." .~:;-~'.""'~~~: '0 'E "u '" Q .",::.', ,-. "- :.'Il ~~;i· "-.~'-""" 0 NI Nil NIII Psae Pybl Pybll Follicular Stago Hybomifra zona/is • .'" llO ~ ~ E 60 ..." '0 'E u "Q '" "- 20 0 NI Nil NIII Psae Pybl Pybll Follicular Stage • • 1~(, Figure 15. Proportions of host-seeking females of obligately autogenous species (i.e. Hybomitra fronta7is, H. hear7ei, and H. pechumani) at successive stages of ovarian development with • abdomens containing four different quantities cf fat body. • H/bornitfél ftontalis '''' 00 01 • 02 lis .. '. > ...... •••••••••••• '. >Cl ...... " ..: ,": o . . .. NI Nil NIII PSélC Pybl Pybll FoIlicular Stogo Hybomilra hootloi "" 00 ....~ E ...... llO 'ô ë ~.. 0 NI Nil NIII Psac Pybl Pybll Follicular Stogo Hybom~ra pechumani 'DO 00 ~ 'a.. E ...... llO 'ô ë .. 0 NI Nil NIII Psac Pybl Pybll Follicular Stago • :::: • APPENOIX 1,),. N:Jrnbcrs of HvbOr.'litr.1 Jnd C~r\,$C\ps spp, dlss~;::('d "':.:rlllc: ~.\ch,\OIL'l'!~ ~\nd day cf the fllgh~ seJsons ct 199D Jnd 1991 Ji CJprlC~rn lrn. 1990 1991 \:cek/OJtc ar .Je :0 1u pc fr he ..lr .le ':0 lu pc fr he olS :1 ni Jul 2 3 la la la 4 la la 5 la 6 la la la 7 la la la la 8 la la 9 la la la la 2 11 12 la la la la la 13 la la .0 14 la la 15 la la la 16 17 la la la 3 18 19 la 20 la la :0 • 21 22 23 la la la 24 la la la la 4 25 la la la la la la 26 la la la la la la la la la 27 la la 28 29 la la la la la la la 30 la la la la la la 31 la la la la la la la la 5 Aug 1 2 la la la la la 3 la la la la la la la la 4 la ------5 6 la la la la la la la la la la la 7 6 8 9 la la la la la la 11 KEY: ar. Hvbomltra arpadi ae • H. ae9uctinct~ zo • H. lonalis lu :II' H. lur;da pe c H. oechumani fr Il H. frontall~ he • H. hearlei a~ Il H. a!:tuta ChrY~oo= zinzalu~ niorlo~~ • z; • ni ;l C. • APPENOIX lb. NUr.lbcr:::. of HI/herni!,",) and Ch~\I~cr~ s'PP. cls:::.c:::.cd cL:r~n~ each wcck and day of the fllght scascns of 1990 and 1991 at Iron A~ fen. 1990 1991 Wcck/O.::tc ar .le zo lu pc fr he ar ole zo , u pc fr he as zi Ju1 1 2 3 la la la 4 5 la la la 6 la la la 7 la la la la la la la 8 la la la 9 la la la la 2 11 12 la la la la 13 la la la la la la la 14 la le la la la la la 15 la la la la la la 15 la la la la la la la la 17 la la la la 3 18 la la la la la la 19 la la la la 20 la la la la la la la • 21 ------_.------._-----.------.- 22 la la 23 la la 24 la la la 4 25 la 26 27 la la la la la la la la la 28 ------.__ .------._--..------.------29 30 la la la la la la 31 5 AU9 1 la la la la la la la la la la 2 la la la 3 la la la la la la 4 la la la la la la -----._------._------._------5 la la 6 la la la 7 6 8 la la la 9 la la la la Il la KEY: ar. Hybomitra arpadi ae • H. aeguetincta %:0 • H. Z'ona11s lu • H. lurida pe • H. peçhu~n\ fr • H. ~~·or.::dl;$ he • H. hearlei as • H. astuta • zi • Chrvsoos zinl~'us • CHAPTER 3. • The Effect of Selected Climatic Factors on the Daily Activity of Tabanids (Diptera) Collected at a Peatland in Subarctic Labrador. • 1. LITERATURE REVIEW 1. Climatic Factors and Tabanid Host-Seeking Activity. The host-seeking activity of female tabanids varies markedly with weather. Among the c1imatic factors correlated with level of tabanid f1ight activity are temperature, light intcnsity, wind, barometric pressure. and relative humidity. 1.1. Temperature. Many authors have noted that there is a temperature below which tabanid host-seeking ceases. The value of this threshold varies among studies, but appears to be around 20°C in most areas of North America (e.g., Blickle 1959, Anderson et aL 1974, Dale and Axtell 1975). Tabanids living in areas where the summer • weather is frequently cold, for exarnple exposed sea coasts (Lane et aL 1983) or subarctic peatlands (Baribeau and Maire 1983, McElligott and Galloway 1991), have reduced temperature thresholds, ranging from 6 to 14°C. Threshold temperature may vary interspecifically at li: given locale, and intraspecifically among populations of a tabanid species at different latitudes (McElligott and Galloway 1991). Host-seeking activity may also be suppressed at temperatures above 35° C (Tashiro and Schwardt 1949, Anderson et aL 1974, Dale and Axtell 1975). Between low and high temperature limits, host-seeking aetivity is highly variable but is generally positively correlated with temperature (Brown and • Morrison 1955, Joyce and Hansens 1968, Schulze et al. 1975, Dale and Axtell 1975, Burnett and Hays 1974. Al\'erson and Noblet 1<17Î. :\urlli and Graf- • Jaccottet 1983.1985, Strickman and H:lgan- 19~D. l\kElligOll. 1<1'lO). 1.2. Light Intcnsity Daily fluctuations in temperature are often strongly correlated with changes in sunlight (Auroi 1978, McElligott 1990). Under certain circllmstances. however, the influence of Iight intensity upon tabanid activity is evidently independent of temperature. Tabanids are generally less active on cloudy days than on sunny ones (Miller 1951, Pechuman 1981), and tabanid f1ight activity has been observed to decline dramatically when clouds obscure the sun (Burnelt and Hays 1974, Leprince et aL 1983). On warm summer evenings. the onset of darkness causes the cessation of tabanid activity long before temperatllres drop • below threshold levels (Hollander 19ii, Anderson 1971. Rober:s 1974. McElligotl and Galloway 1991). 1.3. Wind Speed. Wind affects tabanid f1ight activity, particularly in those areas frequently subjeet to strong prevailing winds. In sea-shore areas (Catts and Olkowski 1972. Joyce and Hansens 1968, Lane et al. 1983), and on hilltops (Leprince et aL 1983), tabanid activity is depressed during periods of high winds. In most areas (e.g., fields, forests), however, wind apparently does not affect tabanid activity (Alverson and Noblet 1977, Strickman and Hagan 1986, McEliigolt and Galloway 1991), • presumably because it does not usually reach high enough speeds; wind speeds 151 ahovc 30 km/hr arc rcquircd bcforc flight is suppressed in coastal tabanid speci", • (Catts and Olkowski 1972. Joyce and Hansens 19liS). lA. Barometric Pressure anù Other Factors. Tabanid flight activity has also been highly correlated with barometric pressure (Alverson and Noblet 1977, Burnett and Hays 1974); horse f1ies are particularly active during periods of low pressure preceding storm fronts (Tashiro and Schwardt 1949, Pechuman et aL 1961). Relative humidity has also been correlated with tabanid flight activity (Bumett and Hays 1974, Dale and Axtell 1975, Alverson and Noblet 1977, Kniepert 1982, Auroi and Graf-Jaccottet 1983), as weil as rainfall, cloud coyer, eV • 1.4. Multiple Regression Analyses. There is considerable disagreement concerning the relative importance of diffcrent climatic factors other than temperature in determining tabanid activity (for review, see Auroi and Graf-Jaccottet 1983). Any attempt to compare various studies is difficult because of differences in species of flies considered, climatic regimes under which studies took place, differences in types of traps and climatic recorders employed, climatic factors considered, and varying degrees of sampling effort. In order to ascertain the effect of one climatic variable on tabanid flight activity \vith any certainty, one must somehow comrol for simultaneously occurring • variables which are orten mutually correlated. One approach which has l'een applied tll this problem is the use nt" • multiple regression analyscs (Burnctt and Hays 1<)74. Schulze CI lIl. \ 'l75. Alyerson and Noblct 1977. Auroi :md Graf-Jaccottct 19S3.\<)S5. Slrickman and I-Iag:tn 1986). Models generated by thesc studies arc usually lirnitcd in predi.:tive value. although they have becn vcry useful in dctcrmining the relative importance of c1imatic factors as they affect tabanid activity. In North American studies. the best multi-species mode!s have bcen ablc to successfully cxplain only about 50Sè of the daily variation in tabanid numbers (Burnelt and Hays 1974. Alverson and Noblet 1977), although Schulze el al. (1975) gcnerated a predictive modd which accounted for 77% of the daily variation in numbers of host-seeking Tabanus nigrovittatus Macquart. Work in Switzerland with Haematopota pluvialis Linnaeus (Auroi and Graf-Jaccoltet 1983.1985), and in Paraguay with Chrysops variegarus • (De Geer)(Strickman and Hagan 1986) has yielded mode\s with R~ values of 0.74 and 0.89 respectively. 2. Interspecilic Variation in Tabanid Host-Seeking Activity. Level of tabanid host-seeking activity has heen found to vary throughout the course of a day independently of c!imatic factors; Kaufrnan and Sorokina (1986) found that the responsiveness of fernale I-Iybomirra spp. to light varies cyclically over the course of a 24 hour period, quite independently of their extemal environment, and that the f1ies' intrinsic, unimodal pattern of "photopreferendum" in the laboratory coincided completely with their rhythm of • daily f1ight activity in nature. Daily activity patterns can vary markcdly among _- 1:' .... tahanid species collected under the same c1imatic conditions (Miller 1951. Harley • 1965, Roberts 1974, Schulzc cr ai. 1975, Burnetl and Hays 1974. Hollander and Wright 1980, McEIEgotl and Galloway 1991). Ali factors considered, it appears that tabanid flight activity is determined on several different levels. At the most basic level, toere is a fly's intrinsic rhythm of daily activity. Next, temperature, light and/or wind thresh01ds determine the period within the fly's period of potential activity during which activity can actually occur. The absolute level of activity, reflected by trap catches, is dctcrmincd by subtle variations in temperature, light, wind, barometric pressure, humidity, and other factors. In addition, the probability of capture of a fly of a given species on a certain day is dependant upon the flight season and peak • abundance of that species; during the peak season, the species' abundance in trap- catches may be truly representative of its activity level, but toward the end of its flight season, there may be few flies around, even though conditions are conducive to flight activity. • • Il. INTRODllCTIO" The host-seeking activity of adult femak horse and deer !lies in northern Canada is confined to the daylight hours (Miller 1951. l'\'IcElIigott and Gallowa)' 1991), and can vary throughout the day in response tll the intrinsie ;lctivity rhythms of the species involved (KauflT'an and Sorokina 19~6) :md the superimposed effects of climatic factors upon flight activity. As with other insects. the flight activity of tabanids may be suppressed if ambient conditions are unfavourable. In subarctic Québec-Labrador, climatic conditions may vary dramatically from one day to the next during the short summer period during which tabanid flight activity occurs (late June-early August). This study wa.~ undertaken in order to (a) accurately establish the temperature and light intensity • thresholds above which female tabanids actively host-seek, and (0) to measure quantitatively the correlations (from half-hour records) of temperature, relative humidity, light intensity, and wind speed on the flight activity of hOSI-sccking tabanids. III. MATERIALS AND METHODS 1. Study Area. This study was carried out during the summer of 1991 at the margin of ,/ • Iron Arm fen, a large peatland located in subaretic Labrador, 20 km northeast of 155 Schefferville, Qut:bec (54°49'N 66 COO'W). This site is described in detail in • Chapter 2 of this thesis. Adult tabanids are known to be abundant at the fen throughout the month of July. 2. Data Collection Methods. A single modified canopy trap was used to monitor the level of tabanid host-seeking activity. Canopy traps selectively collect host-seeking female tabanids. especially those of horse fly species (i.e. Tabanu.~ and Hybomitra spp.); al Iron Arm fen, very few insects other than tabanids were observed in canopy trap collections. The trap was modified by the addition of a touch-sensitive eleclronic "bug counter", (Automata Inc., Grass Valley, Ca.) to the inside of the collecting head of the trap (Fig. 1). This counter was connected to a battery-operated • computerized data logger (Easy Logger™, Omnidata International Inc.. Logan, Utah), together with an air temperature/relative humidity sensor (Omnidata Model ES-UO), a thermistor temperature probe (Omnidata Model ES-060), a Li Cor™ pyranometer (Omnidata Model ES-Z30), and a wind speed and direction sensor (Omnidata Model ES-040). The trap, logger, and ail sensors were placed in the ecotone between the wooded fen margin and the open sedge meadow of the fen. The tempera~J.Ire/relative humidity sensor waS enclosed in an RM Young™ solar radiation shield (Omnidata Model EA-130) mounted on a metal pole at 1.0 m above ground level. At the same height on the pole, the pyranometer was mounted on a leveIling platform, such that its probe was facing • direGlly upwards. The wind speed and direction. sensor was mounted at the top of • the pole. 1.5 m above ground level. Sensors "'t:re plaœd bel"'cen 1 and 2 m above ground levels because this is the lcvcl at which most tabanids probably host-seek (Joyce and Hansens 1968. Schulze cr al. 1975). The data logger \Vas programmed to record the numbt:r of nies collectcd. air temperature, relative humidity, incident solar radiation. \Vinl! speed.•md wind direction at 30 minute intervals from 20:30 hr EDT on 20 July until 19:00 hr on 5 August. Only data obtained bel\veen 05:30 hr and 21:30 hr \Vere analyzed. sincc no tabanids were collected during the hours of darkness. 3. Data Analysis. 3.1. Threshold Values of Temperature and Solar Radiation. • The nightly period of tabanid inactivity at Scllcffcrvi\lc is charactcrizcd bath by darkness and by low temperature; an effon was made ta determine which of these factors regulated the onset and cessation of daily tabanid flight activity. Values for air temperature and incident solar radiation were obtained at the beginning of the 30 minute interval during which the first tabanids of the day were eaptured, and at the end of the 30 minute period du ring which the last tabanid of the day was eaptured. 3.2. Multiple Regression Analyses. Data relating climatic and temporal variables ta number of tabanids • captured during each half-hour interval were analyzed using a multiple regression 15ï • procedure (PROC REG, SAS 1985). In ail regression analyses. date. and half- hour trapping period were considered as discrete variables and weather variables were considered at continuous variables. Those factors of no significance (F test. P:; 0.05) were omitted frorr. subsequent analyses. Multiple correlation coefficients 2 (R ) were computed from regression analyses. A series of regression models were then computed in which weather and temporal factors were deleted one at a time. 2 Multiple correlation coefficients (R ) obtained from the series were then compared with the full-model R2 to indicate the relative influence of each factor on tabanid activity. The greater the influence. the greater the reduction in R2 resulting from the deletion of a single factor. • IV. RESULTS and DISCUSSION In total, fly counts and climatic data were obtained for 509 half-hour intervals, distributed over seventeen 05:30 - 21:30 trapping days. Host-seeking female tabanids were collected, 1 - 38 per interval, during 46% of these intervals. A total of 1167 horse flies and 30 deer flies were trapped. Among the tabanids collected, 64% were Hybomitra arpadi (Szilady), 16% were H. zonalis (Kirby), and 9% were H. aequetincta (Becker). Hybomitra lurida, H. Izearlei, H. frontalis, H. • peclzumani, and H. astuta each made up between 1 and 2% of the flies colleeted; • the remainder was composed of H. affinis. Chr)'sops mer. C. ftm:allls. C. nigripes. C. sordidus, and C. zinzalus. 1. Thresholds. Throughout trapping periods encompassed by the study. air temperature ranged from 2.7 - 27,SOC, relative humidity from 28.6 • 114.6%. solar radiation from 0 - 999.9 W/m2 (Fig. 2), and wind speeds from 0 - 19.2 km/hr. Tabanid activity oceurred only at temperatures greater than 9°C, but acùvity was not an all-or-nothing response to temperature exceeding the threshold value. Rather, at temperatures between 9° and 20°. the percentage of trapping intervals during which tabanids were actually caught increased from 0 to >90%, • according to the linear regression equation (R" = 0.92): y = -78.96 + 8.3SX + e where Y is the percentage of trapping intervals carried out at temperature X in which tabanids were actually collected, and e is the error term (Fig. 3). At temperatures >19°C, probability of capture fluctuates independently of temperature, but generally exceeded 80%. No tabanids were collected at Iron Arm fen on days when the daily maximum temperature recorded at at Schefferville airport was lower than 15°C • (Fig. 4). The airpOrt is located in a relatively exposed location ca. 20 km SW of 15q • lhe sludy area, and air lemperalure lhere on summer days is, on average. 3 - SoC lower lhan at the study area. In general, temperature regulated the onset of activity in the moming, since il was oflen quile sunny by the time thal temperature exceeded the threshold value (Table 1). In the evening, however, il was difficult to separate the effects of temperature and solar radiation, since temperature generally dropped below the threshold value at approximately the same time as night feU. In general, at the cessation of flight activity in the evening temperature was w:;rmer and solar radiation level lower than when tabanid activity began in the moming, suggesting that on sorne days at least, low light, rather than low temperature, was causing tabanid activity to cease for the day. Probability of capture increased wilh • increasing light intensity as weil as with temperature; at higher solar radiation levels, a greater proportion of the intervals trapped yielded flies (Fig. S). 2. Regressions At above-threshold levels of temperature and solar radiation, level of tabanid host-seeking aetivity was determined to a significant degree by, in decreasing order of importance: day of the flight season (p <0.001), temperature (p direction, and wind speed did not appear to be significant determinants of tabanid aetivity under the range of conditions encountered in the present study (p >O.OS). • 'Predictive equations are presented in Table 2. Date, temperature, and lllll • barometric pressure IOgether only accounted for ca. 26':c of tht: diel variation in the activity of Tabanidae. The imponance of date in determining daily aClivily of tab;tnids w;\S mOSI likelya consequence of the timing of this study: tr;tpping w;tS c;\rried out while the overall population of host-seeking tabanids w;\S in decline (Fig. 6). Temperature and relative humidity have been found to be important determinants of tabanid activity in other multiple regression studies (Bumett and Hays 1974, Alverson and Noblet 1977). Another factor reported to significantly affect tabanid activity is barometric pressure (Burnett and Hays 1974, Alverson and Noblcl 1977). Unfortunately, the probe necessary to measure this factor was not available during the present study. • Dominant tabanid species at Iron Arm fen at the time lhat this study wcre Hybomitra arpadi (Szilady), H. zonalis, and H. aequetincta (Becker). Daily activity of H. arpadi and H. zonalis in southeastern Manitoba (McElligott 1990) peaked al approximately the same time of day that overall activity peaked at Schefferville, but the relative contributions of the other Schefferville tabanid species to the overall pattern of daily activity (Fig. 2) are not known, since the data logger approach did not permit identification by species of flies collected during each half-hour period of the day. The incorporation of species specifie data in formulating predictive models is very important; the best predictive models are species specific (Schulze et aL 1975, Auroi and Graf-Jaccottet 1983,1985, • Strickman and Hagan 1986). 161 • 3. The Use of Data Loggers in Future Studies. The results of this study indicate that a computerized data Jogger can be of considerable use in studying the influence of climate on tabanid activity, or the activity of any type of insect which can be attracted to a trap. The logger will carry out accu rate, readings of insect numbers and co·incident readings of a variety of climatic variables (including severa! not measured du ring the present study), and will do so without the necessity of an observer being present. Data loggers can potenùally be placed at remote locations and serviced only on a weekly basis. The use of several loggers or mulùple probes could enable microhabitat-level activity measuremcnt. The main drawback of the use of data loggers in studying insect activity is cost. The unit described in this study costs • approximately S 10,000 Cdn. • • LITERATURE CITED Alverson, D.R., and R. Noblet. 1977. Activity of female Tabanidae (Diptel-ù) in relation to selected meteorological factors in South Carolina. J. Ncd. Entomo7. 14: 197-200. Anderson, J.R., W. Olkowski, and J.B. Hoy. 1974. The response of tabanid species to COz baited insect flight traps in northern California. Pan-Pac. Entomo7. 50: 255-268. Auroi, C. 1978. Les Tabanides (Diptères) de la tourbière du Cachot (Jura neuchâtelois). I. Systématique et méthodes de capture. Bull. Soc. Neuchâte7. Sc. Nat. 101: 27-44. Auroi, C., and M. Graf-Jaccottet. 1983. Influence comparée des facteurs métérologiques sur l'abondonce quotidienne des captures de Haematopota p7uvia7is (L.) et Hybomitra crassicornis Wahlberg (Dipt. Tabanidae) dans de Haut-Jura suisse. Acta Oeco7ogica Gener. 4: 151-165. Auroi, C., and M. Graf-Jaccottet. 1985. Modèle de prédiction du nombre de captures de Haematopota p7uvia7is (L.) (Dipt. Tabanidae) d'après la date et les conditions métérologiques et plaine et en montagne. Acta Oeco7ogica Gener. 6: 179-194 . Baribeau, A., and A. Maire. 1983. Abundance and seasonal distribution of Tabanidae in a temperate and in a subarctic locality of Quebec. Mosq. • News 43: 135-143. Blickle, R.L. 1959. Observations on the hovering and mating of Tabanus bishoppi Stone (Diptera: Tabanidae). Ann. Entomo7. Soc. Amer. 52: 183-190. Burnett, A.M., and K.L. Hays. 1974. Sorne influences of meteorological factors on the flight activity of female horse flies (Diptera:Tabanidae). Environ. Entomo7. 3: 515-521. Catts, E.P., and W. Olkowski. 1972. Biology of Tabanidae (Diptera): mating and feeding behavior of Chrysops fu7iginosus. Environ. Entomo7. 1: 448-453. Dale, W.E., and R C. Axtell. 1975. Flight of the salt marsh Tabanidae (Diptera), Tabanus nigrovittatus, Chrysops at7anticus, and C. fu7iginosus: correlation with temperature, light, moi sture, and wind velocity. J. Med. Entomo L 12: 551-557. Joyce, J.M., Jr., and E.J. Hansens. 1968. The influence of weather on the activity and behavior of greenhead flies, Tabanus nigrovittatus Macquart and Tabanus 7ineo7a Fabricius. J. N.Y. Entomo7. Soc. 76: 72-80. Kaufman, B.Z., and V.V. Sorokina. 1986. Daily rhythms of photopreferendum in horseflies (Diptera: Tabanidae) and their egg parasite Te7onomus angustata • Thomson (Hymenoptera: Scelionidae). Entomo7. Rev. 65: 46-49. 163 Kneipert, F.W. 1982. Der einf1uss verschiedener witterungsfaktoren auf die • flugaktivitat der bremsen (Diptera: Tabanidae). Z. Ang. Entamai. 93: 191-207. Lane, R.S., J.R. Anderson, and C.B. Philip. 1983. Biology of autogenous horse flies native to coastal California: Apatolestes actites (Diptera: Tabanidae). Ann. Entamai. Soc. Amer. 76: 559-571. Leprince, D.J., D.J. Lewis, and J. Parent. 1983. Biology of male tabanids (Diptera) aggregated on a mountain summit in southwestern Quebec. J. Med. Entamai. 20: 608-613. McElligott, P.E. 1989. Seasonal abundance, physiological age, and daily activity of host-seeking horse flies (Diptera: Tabanidae) at Seven Sisters, Manitoba. Unpubl. M.Sc. Thesis, University of Manitoba, Winnipeg, Manitoba. 158. pp. McElligott, P.E., and T.D. Galloway. 1991. Daily activity of horse flies (Diptera: Tabanidae: Hybomitra spp.) in northern and southern Manitoba. Cano Entomol. 123: 371-378. Miller, L.A. 1951. Observations on the bionomics of some northern species of Tabanidae (Diptera). Cano J. Zool. 29: 240-263. Pechuman, L.L., H.J. Teskey, and D.M. Davies 1961. The Tabanidae (Diptera) of • Ontario. Proc. Entamai. Soc. Ontario 91: 77-121. Roberts, R.H. 1974. Diurnal activity of Tabanidae based upon collections in Malaise traps. Mosquito News 34: 220-223. SAS Institute Inc. 1985. SAS users guide: statistics. SAS Institute Inc., Carey, N.C. Schulze, T.L., E.J. Hansens, and J.R. Trout. 1975. Some environmental factors affecting the daily and seasonal movements of the salt marsh greenhead Tabanus nigrovittatus. Environ. Entamai. 4: 965-971. Strickman, D., and D.V. Hagan. 1986. Seasonal and meteorological effects on activity of Chrysops variegatus (Diptera: Tabanidae) in Paraguay. J. Amer. Mosq. Cont. Ass. 2: 212-216. Tashiro, H. and H.H. Schwardt. 1949. The biology of the major species of horse flies in central New York. J. Econ. Entamai. 42: 269-272. • Table 1. Time, tempe rature ('C) and solar radiation (W/mo) at onset and cessation of tabanid activity, Iron Arm fen. July 20 - AU9uSt 5, 1991. ONS ET CESSATION Solar Solar DATE Time Temp . Rad. Time Temp. Rad. July 20 21 • 22 23 11:30 11.8 435.3 19:00 11.1 25.0 24 25 13:00 12.5 272.3 19:30 13.2 45.6 26 9:00 12.5 131.4 19:00 15.3 70.9 27 8:00 15.6 426.3 20:30 15.6 7.9 28 7:00 12.9 144.8 19:00 18.8 68.4 29 8:30 14.1 212.2 20:30 16.9 9.9 30 7:00 16.8 251.6 20:30 17.5 7.2 31 7:30 18.8 310.8 20:00 19.8 6.8 August 1 7:30 13.8 90.1 21:00 17.8 0.1 2 8:30 16.8 359.2 20:30 16.2 5.7 3 6:30 13.8 173.4 21:00 13.3 0.1 4 7:30 18.5 303.8 19:30 18.4 20.6 5 8:30 16.7 270.8 MEAN 14.9 260.1 16.1 22.33 • NOTE: Tabanids were apparently inactive on- July 20 - 22 and July 24. : • • • 165 ., Table 2. Regression equations· for meteorological factors affecting host-seeking activity of adult horse flies at Iron Arm fen, labrador. Factor IXn) b Values for total population Date 0.067" o.on" 0.047" 0.067" Hour of Day -0.002 Relative Humidity -0.038' 0.047" -0.06'" 0.039' Air Temperature 0.217" 0.339" 0.121' 0.214" Solar Radiation 0.000 Wind Direction -0.000 Wind Speed -0.025 --- a Values 0.191 -4.717 7.961 4.630 0.501 ------RZ Values 0.259 0.252 0.240 0.230 0.260 • Prediction equation: Y = a + blXl + bzX z + ... + bnX n + e, where Y = mean tabanid catch/trap/half hour, e = error term, a = Y interccpt, and the b'S determine the slope of the 1ine relating Y to Xn. , p< 0.05, b is significant ., p \1, • Figure 1. Trap he ad modified for use with data logger by addition of • touch-sensitive e1ectronic pad. • • • • RUBBER CORK~ / CLEAR PLASTIC DISK ~ 0 ( MODIFIED CANOPY VAPONA"...mJ '. TRAP HEAD PLASTIC FUNNEL ...... • for use with '.•.... •••.... Data Logger ~ .•..•... ". ~ ..•...... MODIFIED COLLECTING BOTTLE -.. ~ ·······.... ~WIRE SCREEN CONE , .....• ! '..... ; .•..•. '. ~ ..•... TOUCH-8ENSITIVE .....•.. ELECTRONIC PAO ... .J..... ~ '...... ~ REMOVABLE PLASTIC~ ~ ,...... •....•. COLLECTING BOTTLE 1/ METAL SUPPORT RING .... 1. WIRELEAO TO DATA LOGGER-.. VINYL TRAP SKJRT ... • Figure 2. Mean (-), mlnlmum (---), and maximum (---) values of relative humidity, temperature, incident solar radiation, and number of tabanids collected, recorded at half-hour • intervals, 0530-2130 hr EOT, July 20 - August 5, 1991, at Iron Arm fen . • 120 ..--- ... , , 100 ----- ...... _-_ ..... - ...... _...... -- ... .-" • -----_ ... - ... - ... 80 , , ,~ ~ , 60 "'-... , ...... "'--... .. " 40 - ...... -...... ' ---...... ' .. ------_ ..... 20'-'------'---""----...... --'---...... ---'---'---"'- 30 .. -- ...... -- ...... g 25 , _------, -_ .... --- " Cl> , , , ~ :l 20 " ,-- iii~ Q) 15 0. E Q) la 1------... ,------... _----- ... ~ -'- « 5 ------a _1.200 ~ ,ê. 1.000 • ~800 , c: , , 0 , • ==C1l 600 , , ," - ~ '5 , ~ , C1l , , , lI: 400 , , . la ,.' "0 200 , Cf) a 40 • ~ 1 .~ r," ::E 30 ,, " 0 l , ,"' C') ,, , ' :c ,, ,' 1 Cl> 20 , 0. 1... ' 1 , fij- , ...... _... ,,,r,',\ l " , ~ '\ , \' " 1 , , ,- ...... " oi " 1 la , '...... -" ") 1 ~'" ...... \ 1 ) , cr: ,,, , '1 , , '\, " , .' . .,~ • 0730 0930 1130 1330 1530 1730 1930 2130 • Figure 3. Number of half-hour intervals at temperatures ranging from 2 to 27°C, during which tabanids were not collected and during which tabanids were active, as indicated by canopy • trap captures; and percentage of intervals trapped during which tabanids were active, plotted against temperature . • • 40 [:] flios not trappod • me. active ...... ~ ~ ~ .. :;:; .:-: ::: '.' ::: : ::: ..':...... ::;: '.' :::: < : : :: : : : : : : :: : : ---; •••·1··· : : : : :: ::: : : ::: :fi j!: ::: ~.: :.:.:.:.' ::: :: ~~ ~.:.~.:::: ::;: : '::; ::: OL...L:Ju-....w.J-"-'Lt.l.lJ.:.:JLll.Ij;;;.L 2 4 6 8 10 12 14 16 18 20 22 24 26 temperature • _...-go ; 100 1---i.... "--"'''''''' ~:; ; ~__ J c wtic:h fies trIR)Od ~ c: ë Q) Q) .c: ~ ~ c: 20 Q) e y • • 78.85 + B.3:5X Q) R\...... 0. 0 2 4 6 8 ;0 12 14 16 18 20 22 24 26 • temperature • Figure 4. Number of tabanids trapped per day at Iron Arm fen, and daily maximum temperature recorded at Schefferville airport, versus date, June-August, 1990 and 1991. • (Temperature data provided by Environment Canada Climatological Services, A.E.S., 100 Alexis Nihon Blvd., Ville St-Laurent, Qu~bec, H9M 2N8) • 1990 • 2.100 30 '.• . " " 1.800 I~ t l ' ' " ' 1 25 •• , 1 1 :' ,1 , ~ ,•• ,", ",, 1l," :Il:'\ 1.500 . , '1 l , 1 "" Il 1 1 l, \ 1 •, l, 1 • 1" \ • l '. 1 1 • '. • l , , 1 ~ 1.800 . • " " 25 l , " • , 1 ," 1 " , " .. 1\1 • , 1 l , 1.500 " 'I Il'','., ,, , \. ~.. 1 ,'1 _. \,•' .'.' 1., 1 1" Il l "', 20 ci. 1\ l , ,' ''\" r ',\ 1 1 1 o 1 6 11 16 21 26 5 10 15 20 25 30 • June July August • liO Figure 5. Temperature versus incident solar radiation recorded at each of 236 half·hour intervals during which tabanids were • collected, Iron Arm fen, 1991. • •• • 2,000 1 approxlmate tempe~ture threshold , 1 n = 213 observations 1,000 1D0 0 ~O 1 o 6 q:j ~c]J :ftB O:~~ 500 l DO o ocSb 0 0 o C\I DO E I o ~~ ~ o i I@ ~o 8 JRjoo c: gcfE 0 CD 0 o +< 200 I§I ~ .Il! D~~D 0 Do QJo 'etJ o ~ 00 ibP 0lf!J I:!JJ l!§II la 1 ~U cP [EJ S 100 OlPJ 00 cP 0 o o cP @ o r-D 0 0 approxlmale solar 50 1-. 1 0 !Y0 D 0 0 0 ,.radiation Ihreshold ------·---·EJ--r·----·-----·---·g ------O-----.------.--_.-.-.. _----. __ . 1 0 20 1 1 1 1 1 1 5 10 25 30 • 171 Figure 6. Numbers of tabanids collected, June - August 1991 at Iron • Armfen, and period during which present study took place . • • 0 <'l on N CIl 0 -:J N Ol :J ,..on 'Cl:: ,..0 on ,.. <'l CD N ,.. N >. CD "3 ,.. J ,.. • CD CD N ,.. N ID CD C ,.. :J J ,.. CD ,.. o o o o o o o 0 o o o o o o o ,.. et!, &q. CI!. Cl CD <'l CIÏ ,.. ,.. ,.. • _., 1,- • GENERAL CONCLUSIONS 1) The tabanid fauna of the Schefferville ;lrea. as indicatcd by canopy trap and Malaise trap collections of host-seeking adult fcmales in 1990 and 1991. consists of at least 17 species: 6 Clzrysops spp., 10 Hybomitra spp., and one Ary/oms sp.. Hybomitra spp. comprised 96% of collections; Hybomitra arpadi (54%), H, aequetincta (31%), and H. zona/is (7%) were the most commonly collected species. 2) Numbers of adult horse flies and deer flies of different speeies varied markedly between the !wo study sites, Iron ATm fen and Capricorn fen; in general • Iron ATm fen had a more abundant and diverse tabanid fauna than Capricorn. Abundance and diversity of tabanids at the two sites also varied between years of study. 3) Of the tabanid species common in the Schefferville area, H. /urida Fallen, H. aequetincta (Becker), H. arpadi (Szilady), and H. zona/is (Kirby) were most abundant in early and mid July, and H. Izearlei (Philip), and H. peclzumani. were common from mid July until early August. Clzrysops nigripes and C. zinza/us were the most common Chrysops spp.. 4) In Iron Arm fen, a typical tabanid breeding site in the Schefferville area, • larval deer flies (Clzrysops spp.) were collected much more commonly than larva! 173 • horse flies (Hybomirra spp.). Of the 476 tabanid larvae collected, 82.7% were Chrysops (5 spp.). 17.0% were Hybomirra (5 spp.), and 0.3% were Arylotus shagnicola. The most abundant species in the fen w::-re C. =in:alus (31 %). and C. nigripes (24%). 5) Species- and genera-specifie microhabitat preferences were apparent in tabanid larvae; in general Chrysops spp. preferred drier regions of the fen than did Hybomitra spp.. 6) Clzrysops zinzaius and C. nigripes appear to require 3-4 years to complete their life cycles in subarctic regions, based on seasonal larval growth patterns. Annual • variation in larval year-class sizes suggests that populations vary dramatically within individual peatlands from year to year; this is supported by the high degree of annual and local variation apparent in numbers of adult tabanids. 7) In the Schefferville area, H. arpadi and H. aequetincra are obligately anaulOgenous (i.e., require a blood·meal in order to mature eggs), H. iurida and H. zonaiis are facultatively aUlOgenous (i.e., capable of. maturing eggs without in sorne cases), and H. pecllumani, H. Izeariei, H. frontaiis (Walker), H. astuta (Osten Sacken), C. zinzaius and C. nigripes are obligately autogenous (i.e. always mature a first batch of eggs without taking blood). • 1-,.., • 8) Based upon gonotrophic age-graùing of nul1ipamu,; individua!,;. th.: majllrity of H. aequetincta and H. arpadi femalc,; emerge either at the beginning of the flight season, midway through the season, or bath. depending upon ye:lr :llld ,;ite. Most H. zonalis emerge mid\Vay through the flight ,;ea,;on. 9) Nulliparous female tabanids of anautogenous or facult:ltively autogenolls species usually carry considerable amounts of fat body \Vithin their abdomens. This fat is depleted as nutrients are transferred to the developing oocytes; reccntly parous females carry very little fat body. 10) Host seeking activity by female tabanids only occurred at temperatures 2 • exceeding 9° C and levels of solar radiation exceeding 0 W1m • The onset of tabanid activity in the morning \Vas uSllal1y temperature dependant, whereas the cessation of activity in the evening \Vas light dependent. Tabanid activity increa.~es \vith increasing temperature between 10 and 19° C. 11) During the daylight hours, temperatllre, relative humidity, and date of flighl season most significantly affecl level of tabanid hast-seeking activity; host-seeking is not signifieantly affected by Hour of day, solar radiation, wind direction, or wind speed, under the conditions encountered during period when the sludy wa.~ condueted. • .. 175 Claim To Originality: The following are original contributions to knowlege: 1) A detailed description of the faunal diversity of larval and adult Tabanidae in peatlands in the vicinity of Schefferville, Quebec. 2) A description of the rnicrohabitat preferences of larval tabanids in a subarctic peatland. 3) A description of the annual growth patterns and annual variation in the size of cohorts of larval Chrysops nigripes Zetterstedt and C. zinzalus Philip. 4) An illustration of the larva presumed ta be that of C. zinzalus. 5) A descriptior:: of the degree to which faunal diversity of adult tabanids varies locally, seasonally, and annually in the Schefferville area. 6) A description of the seascnal variation in reproductive age of tabanids in • subarctic Canada. 7) The use of a reproductive age-grading technique to infer emergence patterns of anautogenous and facultatively autogenous tabanid species. 8) The observation that populations of H. lurida (Szilady) and H. zonalis (Kirby) are facultatively autogenous, and that H. astuta (Osten Sacken), C. nigripes, and C. zinzalus are autogenous species. at least in the Schefferville area. 9) Determination of fat body depletion and accumulation patterns of subarctic tabanid species. lOr'The use of a multiple regression model to determine the influence of , meteorological variables on the activity of tabanids in the subarctic. •