Publications in Publications de Natural Sciences sciences naturelles QH «| 7 - I B82. NaF iv National Museums Musées nationaux of Canada du Canada National Museum Musée national of Natural Sciences des sciences naturelles 3 An Analysis of Toads of the Bufo americanus Group in a Contact Zone in Central Northern North America Francis R. Cook nici ee ~~ Ee UAL! ORNIA y a] Af i rv\rn ay P - = 'q H} ACADEMY OF SClENC ee |3 J ACADEMY OF SCIENCES | . MAY | 4 ) ia | i Hy LIC XARY t en ean if “ve ai? ‘ <" : nr > National Museums National Museum of Canada Ottawa 1983 Publications in of Natural Sciences Natural Sciences, No. 3 AN ANALYSIS OF TOADS OF THE BUFO AMERICANUS GROUP IN A CONTACT ZONE IN CENTRAL NORTHERN NORTH AMERICA Francis R. Cook Herpetology Section National Museum of Natural Sciences Ottawa, Canada K1A OM8 Publications de sciences naturelles, n° 3 Musées nationaux Musée national du Canada des sciences naturelles Publications in Natural Sciences, no. 2 Publications de sciences naturelles, n° 2 Coad, Brian W., H.F. Alkahem, and R.J. Behnke. Acanthobrama hadiyahensis, a new species of Cyprinid fish from Saudi Arabia. Published on May 24, 1983. National Museum of Natural Sciences Publications in Natural Sciences, No. 3 Published by the National Museums of Canada ©National Museums of Canada 1983 National Museum of Natural Sciences National Museums of Canada Ottawa, Canada Catalogue No. NM. 95-43/3E Printed in Canada ISBN 0-662-11274-1 ISSN 0714-0983 Musée national des sciences naturelles Publications de sciences naturelles, n° 3 Publié par les Musées nationaux du Canada ©Musées nationaux du Canada 1983 Musée national des sciences naturelles Musées nationaux du Canada Ottawa, Canada N° de catalogue NM. 95-43/3E Imprimé au Canada ISBN 0-662-11274-1 ISSN 0714-0983 Contents Abstract, v Résumé, v1 Acknowledgements, vii Introduction, 1 Materials and Methods, 3 Description of areas sampled, 3 Field surveys, 4 Morphological character set, 5 Call analysis, 12 Computer analysis of data, 13 Hybridization experiments, 14 Results, 15 1. Field surveys, 15 2. Analysis of morphological variation for reference samples of B. a. americanus and B. a. hemiophrys, 16 a) Discriminant analysis of males, 16 b) Univariate and ratio analysis of males, 20 c) Discriminant analysis of females, 28 3. Morphological character comparison of southeastern Manitoba Bufo, 28 a) Discriminant analysis of males, 28 b) Univariate and ratio analysis of males, 36 c) Discriminant analysis of females, 43 4. Breeding call analysis and comparisons, 44 5. Crosses reared in laboratory and field, 51 6. Comparisons of hemiophrys and americanus with other Bufo in northern North America, 52 a) Hemiophrys - boreas, 52 b) Hemiophrys - cognatus, 55 c) Hemiophrys - w. woodhousei and americanus - w. fowleri, 56 Discussion, 57 a, Habitat, 57 a) Breeding habitat, 57 b) Summer habitat, 57 2. Morphology, 57 a) Non-quantitative, 57 b) Quantitative, 57 c) Selective advantage of taxonomic characters, 60 Breeding call analysis and comparisons, 61 Cross-breeding experiments, 63 5. Taxonomic conclusions, 63 a) Comparisons of B. a. hemiophrys and B. a. americanus with other Bufo spp. in northern North America, 63 b) Status of B. americanus copei and B. hemiophrys baxteri, 65 c) Status of B. americanus and B. hemiophrys, 66 6. Zoogeographic significance of the Eastern Manitoba transition zone, 69 7. Taxonomic significance of. the americanus - hemophrys relationship with respect to the other nominal species of the B. americanus group, 71 WwW ill References, 74 Appendix I: List of collections used in morphological analysis, 77 Appendix II: List of collections used in call analysis, 87 Submitted for publication: October 5, 1978. Accepted for publication: January 28, 1983. Abstract Discriminant function analysis of 30 measured morphological characters was used to compare reference samples from Alberta to Prince Edward Island of 825 male Bufo a. americanus and 819 male B. a. hemiophrys and 59 female B. a. americanus and 118 female B. a. hemiophrys. Both analyses provided complete separation of the two taxa. When the resulting weights were applied to 2460 male and 225 female Bufo from southeastern Manitoba, an area excluded from reference samples, a relatively narrow zone of in- termediate populations was defined. Separate analysis of three additional scored characters and of selected measured characters, alone or as ratios, followed the same pattern. Audiospectrograms of male breeding calls substantiated the presence of the intermediate populations and laboratory- and field-raised crosses between B. a. americanus, intermediate, and B. a. hemiophrys parents in- dicated some transformation success in all combinations. | Discriminant scores for B. woodhousei fowleri (19 males) and B. w. woodhousei (17 males) gave mean values different from the closest B. a. americanus or B.a. hemiophrys populations. Separate discriminant analysis comparing B. a. hemiophrys with B. cognatus (48 males) and B. boreas (60 males) showed complete separation of the taxa in areas of sympatry, and the mor- phological intermediacy of a single B. a. hemio- Dhrys x B. boreas field-identified hybrid was confirmed. The taxa B. a. americanus and B. a. hemiophrys are conspecific and lack apparent barriers to inter- breeding with introgression indicated east and west of a narrow primary contact zone. Because of the narrowness of the primary zone of intermediate populations (about 16 km) and the distinctiveness of these taxa outside of this zone, they can be con- sidered megasubspecies: B. (americanus) americanus and B. (americanus) hemiophrys. Two previously recognized subspecies, B. a. copei (northern Ontario and Quebec) and B. h. baxteri (Wyoming) are not here regarded as taxonomically distinct. The intermediate zone between B. a. americanus and B. a. hemiophrys corresponds to the mid- point in transition between eastern and central herpetofaunas, elements of which are postulated to have been separated during a break in the trans- continental Boreal Forest during the Wisconsin, and perhaps earlier, glaciations. Résumé On a procédé a des analyses au moyen de fonc- tions discriminantes de 30 caractéres morphologi- ques pour comparer des échantillons provenant de |’ Alberta a l’ile du Prince-Edouard et comprenant 825 Bufo a. americanus males, 819 Bufo a. hemiophrys males, 59 B. a. americanus femelles et 118 B. a. hemiophrys femelles. Les deux analyses ont révélé une séparation complete de ces deux taxons. L’application des résultats pondérés obtenus chez 2460 males et a 225 femelles Bufo du sud-est du Manitoba, une région non représen- tée dans les échantillons de référence, a permis de déterminer |’existence d’une zone relativement étroite ou se trouve des populations intermé- diaires. Une analyse séparée de trois autres carac- teres et d’une sélection de caractéres mesurés isolé- ment, ou sous forme de rapports, a été effectuée suivant le méme schéma. Des audiospectrogram- mes du chant nuptial d’un certain nombre de males ont confirmé la présence de populations intermédiaires, et l’élevage en laboratoire et dans la nature des produits du croisement de parents B. a. americanus intermédiaires et de B. a. hemiophrys a abouti a la métamorphose complete de spécimens issus de toutes les combinaisons. Les résultats discriminants concernant B. wood- housei fowleri (19 males) et B. w. woodhousei (17 males) reflétaient des valeurs moyennes différentes de celles qu’a livrées l’analyse des populations les plus étroitement apparentées de B. a. americanus vi et de B. a. hemiophrys. Une analyse discriminante distincte visant 4a comparer des populations de B. a. hemiophrys a des populations de B. cognatus (48 males) et de B. boreas (60 males) a révélé une séparation complete des taxons dans les zones de sympatrie et a confirmé la position intermédiare d’un unique hybride de B. a. hemiophrys et de B. boreas qui avait été identifié sur le terrain. Les taxons B. a. americanus et B. a. hemiophrys appartiennent a la méme espéce et rien ne semble empécher I|’hybridation. Une introgression a été signalée a l’est et a l’ouest d’une étroite zone de contact primaire. Etant donné la faible étendue de l’habitat principal des populations intermé diaires (environ 16 km) et le caractére distinctif de ces taxons a l’extérieur de cette zone, ils peuvent étre considérés comme des mega-sous-espéces: B. (americanus) americanus et B. (americanus) hemiophrys. Deux autres sous-especes reconnues antérieurement, B. a. copei (nord de |’Ontario et du Québec) et B. h. baxteri (Wyoming) ne sont pas distinctes du point de vue taxinomique. La zone intermédiaire entre les populations de B. a. americanus et de B. a. hemiophrys corres- pond au centre de la zone de transition entre les faunes herpétologiques de l’est et du centre, dont les éléments auraient été isolés par une interrup- tion de la forét boréale transcontinentale lors de la glaciation du Wisconsin, ou peut-étre lors d’une glaciation antérieure. Acknowledgements Joyce Crosby Cook actively participated in all field surveys, from 1963 to 1970, encouraged the work throughout, and read the entire manuscript. W.J. Cook and T.W. Cook have also aided various stages of this project. J.S. Bleakney was Curator of Herpetology at the National Museum during 1952-1958, and reference samples from this period as well as from Prince Edward Island in 1959 were made by him alone (1952-1953) or with my participation. Richard Otterdahl (1959), R.A. Henry (1960), M. Gordon Foster (1965), Stanley W. Gorham (1965), and C. Bruce Powell (1962, 1963, 1965) assisted my field programs which collected much of the re- maining reference material. K.W. Stewart, Tom Vincent, Jim Johnston, Patrick Gregory, G. Viscotto, R. Merkel and J. Gilhen participated in portions of the southeastern Manitoba surveys. Andrew Kozlowski, F.W. Schueler, Wayne F. Weller, Ross MacCulloch, D.S. Redford and D. Summer made special efforts in obtaining addi- tional reference collections used in this study. Ken- neth R. Porter graciously made available his samples from the Laramie Valley. K.W. Stewart, Zoology Department, Universi- ty of Manitoba, deserves special thanks for his un- bounded enthusiasm, advice, patience and criticism. Frederick W. Schueler ran the major portion of the morphological data on the University of Toronto computer, contributed statistical insights and advice throughout, and searched out the geographic coordinates. Don E. McAllister, Curator of Ichthyology, National Museum of Natural Sciences, made the ‘‘ICHTHOS’’ com- puting facility available and also advised on statistical problems, as did Roger Green and Douglas Copeman. Wendy Antoine, then of the Canadian Aquatic Identification Centre of the NMNS, converted a sea of crowded pencil notations into essentially flawless computer cards. D. Faber, L. Marhue and J. Fournier made this contribution possible. Charlie Douglas, staff artist at the National Museum, painstakingly produced the views of Bufo included here. Vii Jim Johnston, as Curatorial Assistant in Herpetology from 1972, catalogued specimens, produced audiospectrograms from the original tapes, and recorded the sound variables used from them, ran data on the ‘‘ICHTHOS”’ computer, drafted and photographed the graphs and maps included here, and aided in many other ways. Marilyn Fisher typed difficult first drafts of the manuscript, and Jan McConnell and her staff at the National Museums’ Word Processing Centre the several subsequent manuscripts, and Thérése Lapierre made final corrections. Eleanor Fenton saw the manuscript through to publication, and Bonnie Livingstone provided her expertise on for- mat. The Translation Bureau, Secretary of State, prepared the translation of the Abstract. This study was supported and funded throughout by the National Museum of Natural Sciences, and my appreciation and gratitude is acknowledged to L.S. Russell, A.W.F. Banfield, L. Lemieux, E.L. Bousfield, F.H. Schultz, T.A. Russell, W.E. Godfrey and H. Ouellet for mak- ing facilities available and supporting field work, and leave at the University of Manitoba, and pro- viding the time necessary for this work. The University of Manitoba provided facilities during 1968-1970, and particular thanks are due to Harold E. Welch and K.W. Stewart for these. Na- tional Research Council grants to K.W. Stewart, University of Manitoba, and J. Rising, University of Toronto, supported certain equipment (rearing cages and pans and computing time, respectively) for this study. D.E. McAllister, C. Gruchy and F.W. Schueler read earlier drafts of the manuscript and made many useful comments and suggestions; K.W. Stewart, C.C. Lindsey, Jack Gee of the Zoology Department, R.E. Longton of the Botany Depart- ment, University of Manitoba, and J.P. Bogart of the University of Guelph, read and criticized the entire manuscript as submitted to the Univer- sity of Manitoba in partial fulfilment of the re- quirements of the degree of Doctor of Philosophy (Cook, 1978) and R.G. Zweifel, American Museum of Natural History commented on a later draft. The bulk of their suggestions have been in- corporated into the present revisions to its benefit but there remain areas of interpretation, par- ticularly in taxonomy, where one or several of these reviewers would have strongly preferred a different conclusion. The retention here of my original conclusions implies no lack of awareness, respect or appreciation for their arguments. The late E.B. Shelley Logier, formerly Curator of Herpetology of the Royal Ontario Museum, first stressed to me the lack of herpetological knowledge of central Canada and the need for in- tensive studies there nearly 40 years ago. This work is a direct result of those conversations. Introduction Five species of toads of the 187 (Gorham 1974) recognized in the nearly cosmopolitan genus Bufo (Amphibia: Anura: Bufonidae) have been recognized in northern North America (Conant 1975; Stebbins 1966). Three are broadly distributed: the American Toad, Bufo americanus, the Canadian Toad, B. hemiophrys, and the Northwestern Toad, B. boreas. These are largely allopatric and replace each other from east to west (Figure 1). North-south boundaries lie through eastern Manitoba and western Alberta. All are widespread in the boreal or montane forests which cover much of this region, but none invade tun- dra or true alpine conditions. They are abundant in the deciduous forests of the east (B. americanus), the parkland and northern fringes of the grassland of the central portion (B. hemiophrys) and the arid interior valleys and coastal rainforest of the west (B. boreas). Two ad- ditional species invade or approach the southern fringes of the region. In the east, Fowler’s Toad, B. woodhousei fowleri, reaches its northern limit along the sandy north shore of Lake Erie in southern Ontario. Here and south to the Atlantic coastal plain it is geographically sympatric with, but largely ecologically separated from, Bufo americanus. In the central portion of the conti- nent Woodhouse’s Toad, B. woodhousei woodhousei, is allopatric to Bufo hemiophrys along a line from northern Montana to northern South Dakota (Figure i). The Great Plains Toad, B. cognatus, which is widely sympatric with B. woodhousei over the interior of the continent south to northern Mexico, is narrowly sympatric with B. hemiophrys along the grassland southern fringe of its range in southeastern Alberta and southwestern Saskatchewan (Figure 1). A disjunct population of B. hemiophrys occurs in the Laramie Valley of Wyoming. In western Alberta and in British Columbia, the Yukon and coastal Alaska, B. boreas is the only toad, but to the south its range interdigitates with other species (e.g. B. w. woodhousei in Washington and further south). The literature on toads is voluminous (Blair 1972) and the evolutionary relationships, and many of the comparative aspects between species are extensively documented (osteology, chromo- somes, electrophoretic patterns of proteins, venom constituents, and laboratory hybridization, among others). Blair (1972 and previous papers) has placed all the species considered here in his narrow-skulled (or thin-skulled) northern evolu- tionary line within the genus, and concluded that they represent three species-groups. B. americanus, B. hemiophrys and B. woodhousei are all included in the americanus group; B. cognatus and B. boreas each are placed in distinct groups. However, relatively little has been pub- lished on northern populations of these taxa, and their geographic variation and interactions within this region have been poorly documented or are unknown. This study attempts to clarify the relationship and interaction between B. americanus and B. hemiophrys. It has been possible to examine a larger number of specimens than any previous authors had available from over the combined range of the two forms. This sample encompasses northern North America from Prince Edward Island to the eastern foothills of the Rocky Moun- tains in Alberta. The geographic variation of their populations over this vast area is compared to the variation in population samples obtained through transects of their contact zone in eastern Manitoba. Calls of individuals from intermediate populations are compared to those of both eastern (B. americanus) and western (B. hemiophrys) populations adjacent to the contact zone and the relationship of morphological variation and call variation is compared. Data on crossing success are less extensive and less satisfactory but it is the first attempt to field-rear tadpoles of crosses be- tween these forms and compare their viability in different pond environments. Also discussed are comparisons between the B. americanus-hemiophrys complex and samples of B. w. woodhousei, B. w. fowleri, B. cognatus and B. boreas. Throughout the remainder of this paper I an- ticipate its conclusions and refer to the taxa previously called Bufo americanus and Bufo hemiophrys as B. a. americanus and B. a. hemiophrys. Figure 1. The geographic distribution of toads in northern North America based on Conant (1975) and Stebbins (1966), with modifica- tions and additions from National Museum of Natural Sciences (Herpetology Section) collections and Cook (1977). : 5 iM i ying 4 id ; Pia ‘ & ( rs. i wens \ 50 Px % ed L 7 VAR Sat bel ~ Fy ss iy V9) ; R on ocala a maa Ce me a ao a ee Bufo a. americanus, B. a. hemiophrys and B. boreas replace each other across the continent from east to west, respectively. Note the narrow zone of sympatry between B. a. hemiophrys and B. boreas in central Alberta, and the disjunct population of B. hemiophrys in Wyoming (bottom of map, west of centre). The Bufo woodhousei complex: B. w. woodhousei in the west and B. w. fowleri in the east. Compare with the top map for area of sympatry between B. w. fowleri and B. a. americanus in the east, for allopatry between B. w. woodhousei and B. a. hemiophrys in the central portion and sympatry between B. w. woodhousei and B. boreas in the west. Ecological separation occurs within the areas of geographic overlap. Bufo cognatus. Compare with top and middle for areas of sympatry with B. a. hemiophrys and B. w. woodhousei. There is some ecological separation between B. cognatus and the other species within their overlapping ranges. Materials and Methods Description of Areas Sampled The primary vegetation zones of Canada have been discussed and mapped by Rowe (1959) and these are used on the base map showing position of reference samples of Bufo a. americanus and B. a. hemiophrys in eastern and central North America (Figure 2). Forest region distributions depend on a variety of climatic, topographic, soil, and geologic variables and are, at least in part, a function of the combined effect of these on gross vegetation. Amphibian and reptile distributions are influenced by the same variables and often show some general agreement with vegetation zones (Savage 1960) though exact range limits for particular species vary as much as those of individual tree species. Forest regions are used here as a crude measure ] j *o°o° 1 Great Lakes =] Western Parkland Prairie of climatic and other variables which affect toad distribution. Southeastern Manitoba was most intensively sampled because it includes the contact zone be- tween toad taxa that is central to this study. Figure 3 shows the collecting localities and vegetation types in this area. The southern tongue of the Manitoba lowlands section of Rowe (1959) is im- portant in the following analysis. This is an area of change from predominantly coniferous trees of the Boreal and Great Lakes complex to aspen and grassland vegetation. At some localities (e.g. the Brokenhead River at Highway 1; Locality 35) the change is abrupt and can be readily appreciated in the field. However, various coniferous outliers exist to the west (e.g. at Birds Hill), and simple correlations of toad morphology and tree species cannot be readily made. 9 oo © o,°0 e@e@e @ @i-nA*/@ @ @ Subalpine = Montane ate Eastern Deciduous Northern Desert , South-Eastern Lo Saad onrea udy Oak -Savannah tained. These localities are numbered separately within each province or state and are listed in Appendix I with catalogue numbers for each collection examined. The forest regions shown are simplified from Rowe (1959) with southern extensions from Clarke (1973). Grassland Aspen — Oak Manitoba Lowlands Great Lakes % oo Ds Figure 3. Map of southeastern Manitoba showing the localities from which collections of toads for analysis were obtained. Local- ity numbers are listed in Appendix I with catalogue numbers of specimens examined. The forest regions shown are simplified from Rowe (1959). The dashed line represents the mid-point of the contact zone between Bufo a. hemiophrys (west of line) and B. a. americanus (east of line) as discussed in the text. Southeastern Manitoba has been modified by clearing for tillage and drainage for agriculture, primarily in its western portion (Warkentin 1967). The vegetational transition and area of contact between toad taxa features abundant beach ridges of glacial Lake Agassiz and is bordered on the east by peat and swamp and to the west by clay soils (Figure 4 in Elson 1967). Love (1959) and Shay (1967) have presented discussions of the post-glacial vegetation in this area, demonstrating the changes that have oc- curred over the past 12 000 or more years. Field Surveys Field work was conducted largely by motor vehicle along highways and rural roads. For this reason roadside ditches, ponds, lakes, and rivers were the main areas sampled. Breeding toads were of prime concern. These are here assumed to provide the best population samples for geographic com- parisons because the individuals are known to be mature adults. However, additional samples not included here were also taken throughout the ac- tive season for amphibians and reptiles whenever possible, in order to obtain tadpoles of various growth stages, metamorphosing individuals, and samples of juveniles and adults outside the reproductive season. General survey collecting sites were chosen to provide a wide geographic sample. In Manitoba, transects were run across the southeastern corner of the province on the east-west network of public roads. Particular attention was given to intensive sampling where the putative zone of taxa contact seemed most obvious. An eight-cell headlamp was used for illumina- tion during after-dark collecting. At peak breeding, and usually at other times as well, this artificial light is apparently ignored by toads and they continue normal activity. Attempts were made to obtain a minimum of 20 breeding males per locality, although this was not always possible. Much larger series (100 or more) were occasionally taken to provide a more robust sample for statistical analysis. Many Trans- Canada Highway localities were resampled to allow analysis for possible within- and between- year variation. Generally, a large proportion of males utilizing a particular site for breeding could be expected to be present on a peak breeding night, but females tended to arrive over a more extended period though the breeding season, breed shortly after joining the chorus, and leave after depositing their eggs. This results in a few females relative to the numbers of males being available at any sampling period. In addition, females tend to come to the ponds in greatest numbers on a few peak nights of the breeding season, whereas males, though in reduced numbers compared to peak intensity, con- tinue to call and be present over much of the breeding season. Because of their call, males are also more conspicuous. Juveniles were rarely pre- sent at the breeding site, but when present could be recognized as immature by size and lack of breeding condition. Field notes on habitat, general topography and weather conditions, as well as relative density and behaviour of toads, were made for each locality at the end of each sampling period. Specimens for preservation were usually pro- cessed within 12 hours of capture, but occasionally they were retained in dampened cloth bags under refrigeration for longer periods. All were killed with ether, were injected with 10% formalin and individually tagged, then im- mersed in formalin of the same strength (Cook 1965a). Permanent museum catalogue numbers were assigned to each collection in the field, and individuals were given a subnumber within that catalogue number. For musuem storage, collec- tions were transferred to 45% isopropyl alcohol. Morphological Character Set The following character set was devised primarily to compare B. a. americanus and B. a. hemiophrys on the basis of characters which appeared to dif- fer between the two taxa or which showed varia- tion of interest within each taxon. It was later ap- plied to one sample of B. boreas and one of B. cognatus and to small samples of B. w. woodhousei and B. w. fowleri to compare their variation in these characters with B. a. hemiophrys and B. a. americanus. Additional characters could have been added which would have improved the separation in certain of these combinations. Many of these characters, in the same or slightly different forms, have been used previously to dif- ferentiate B. a. americanus and B. a. hemiophrys. Blair (1957a), Wright and Wright (1949), Brecken- ridge (1944), Conant (1975) and others have presented comparative descriptions. Previous at- tempts at quantification have been made by Underhill (1961), to compare variation within populations of B. hemiophrys and between B. hemiophrys and B. woodhousei and B. ameri- canus. Porter (1968) compared the Wyoming relict with samples from the continuous range of B. hemiophrys, and Henrich (1968) assessed popula- tions from South Dakota which he showed to be intermediate between B. hemiophrys and B. americanus. All measurements and coding states used in the present analysis were made after at least two years preservation. Every measurement was taken with the same model dial calipers to the nearest 0.1 mm by myself over a period of four years, and geographic sequence of measuring collections was random. Lee (1982) has presented a statistical analysis showing significant variation in the measurements of Bufo marinus taken before preservation and after preservation periods of four and 14 months. Based on these results he questioned the validity of comparisons in morphometric studies that are based on preserved material. In my study the variation between collections is minimized by most collections having been processed similarly by J.S. Bleakney or myself (the most notable exception is the RKP Wyoming collection which was stored in ethyl rather than isopropyl alcohol) though the relative periods in formalin or alcohol did vary by unrecorded amounts. The collections may have been in isopropyl alcohol long enough to have stabilized, and the general consistency of the results seems to indicate that preservation and measurement variations over time do not in- validate the analysis. Fresh measurements that I had taken for all specimens that I collected per- sonally in the field were deliberately not used in the present study because of the obvious problem of shrinkage of individuals after preservation and the need to measure additional characters and to compare material for which fresh measurements by the same investigator were not available. Lee’s (1982) point that preserved specimen measure- ments do not exactly represent the size or propor- tions of a living animal are valid and the caution inherent in his paper about overinterpretation of small differences has hopefully been intuitively avoided throughout the discussions presented here. The following measurements were recorded for all specimens except for some Bufo boreas in which the cranial crests were not obvious, thus for- cing the omission of all characters involving these structures. Figure 4 depicts the measured characters and Figure 5 the ventral pattern variation. Figures 6 and 7 compare dorsal views of toads studied, and Figure 8 depicts ventral views of hind feet. Figure 4. Diagrammatic views of a generalized toad showing underside of right hind foot, dorsal view of head and shoulders, and dorsal and lateral aspects of a whole animal. CA: cranial crest: anterior width CL: cranial crest length CP: cranial crest: posterior width EL: eyelid length HW: head width IW: tubercle width (inner metatarsal tubercle) LW: diameter of largest wart on tibia NP: nostril to parotoid NS: nostril separation OW: ‘‘spade’’ (outer metatarsal tubercle) width PL: parotoid gland: length PS: parotoid gland separation PW: parotoid gland: width T: tibia length TH: tympanum height TL: tarsus length SL: spot length SV: snout-vent length SW: spot width WW: wart width Figure 5. Diagrammatic venters of toads showing the range of ventral spotting and the scoring values (0-6) assigned (see text for definitions of scored conditions). Figure 6. Dorsal views of selected male toads in the Bufo a. americanus-B. a. hemiophrys complex showing external structure and pattern. A. B. a. americanus, NMC 14797-4, Whitetop Creek, Moosonee, Ontario . a. americanus, NMC 11795-3, 3 miles [4.8 km] west on Highway 1 of junction Highways 1 and 11, Manitoba . a. americanus x hemiophrys, NMC 8484-11, 17 miles [27.4 km] west on Highway 1 of junction Highways | and 11, Manitoba . a. hemiophrys, NMC 12217-1, 30 miles [48.3 km] west on Highway 1 of junction Highways 1 and 11, Manitoba . a. hemiophrys, NMC 8543-2, 10.5 miles [16.9 km] west and north on Highway 28 of St. Paul, Alberta moaw Baw Figure 7. Dorsal views of other Bufo taxa and one hybrid. . B. boreas, NMC 17014, Mile 1.9 [Kilometre 3.1] on Kananaskis Lakes Road, Alberta B. boreas x B. a. hemiophrys, NMC 8523, 2.4 miles [3.9 km] north on Highway 44 of Westlock, Alberta . B. cognatus, NMC 4168, 3.1 miles [5.0 km] east on Highway 1 of junction Highways 1 and 21, Saskatchewan . B. w. woodhousei, NMC 8591-6, 3.5 miles [5.6 km] SSE of Stockton, Rooks County, Kansas . B. w. fowleri, NMC 15972-10, Long Point, Norfolk County, Ontario moaw> Figure 8. Ventral views of the right hind foot of selected toads. . B. a. hemiophrys, NMC 8521-35, 2.4 miles [3.9 km] north on Highway 44 of Westlock, Alberta. B. a. americanus, NMC 7064-33, 10.5 miles [16.9 km] east on Highway 17 of North Bay, Ontario. B. boreas, NMC 8530-33, 2 miles [3.2 km] west on Highway 2 of Athabaska, Alberta. . B. cognatus, NMC 7110-5, 8.3 miles [13.4 km] east on Highway 1 of Piapot (turnoff), Saskatchewan. B. w. fowleri, NMC 15783-55, Long Point, Norfolk County, Ontario. mow > —" . Sex: specimens were sexed by presence of nuptial pads and darkly pigmented vocal sac in males, both of which are absent in females. Since most collections analysed were from breeding populations this was generally adequate. Specimens less than the breeding size of the smallest mature male for that locality were designated as juveniles, and omitted from analyses. Snout-vent: taken from the tip of the snout to the centre of vent, with the toad pressed as flat as conditions of preserva- tion would allow. Nostril separation: the distance separating the nostrils measured across the nose be- tween their interior (medial) edges. “Cranial crest’’ length, right and left sides: measured from the nostril to the posterior edge of the crest. The anterior edge of the cranial crest merges with the bony structure of the nose so variously that the nostril was the best fixed point for anterior reference although this necessitates including more than the crest itself. Cranial crest type: four arbitary catego- ries were based on the difference between extreme B. a. americanus and B. a. hemiophrys and the intermediate condition. 0) - a solid, filled-in boss caused by the cranial crests lacking a groove be- tween them or having only a slight trace of a groove. The extreme B. a. hemiophrys condition. 1) - crests parallel or divergent anterior- ly with a bridge joining them across the back, and a distinct, but shallow, groove between them in the boss. Many B. a. hemiophrys. 2) - crests variable, but always with either a trace of filling-in between them or a partial posterior bridge. Common in many individuals from hybrid populations. 3) - crests distinct, no trace of filling-in between them and usually con- vergent anteriorly and strongly divergent posteriorly, but often hooked in slightly toward each other dq; 8. at the extreme posterior. The typical B. a. americanus condition. Of the other toads analysed, B. cognatus and B. woodhousei were also scored in this category. The anterior con- vergence and posterior divergence is extreme in B. cognatus, and the crests usually less divergent in B. woodhousei. B. boreas has low in- distinct crests if they are visible at all, and all crest characters were omitted for this species. Nostril to parotoid, right and left sides: the distance from the middle of the nostril to the anterior edge of the parotoid gland on both the right and left sides. A prob- lem sometimes arose in defining the anterior edge of the gland, which may be somewhat indistinct especially in cases where it was pressed against the post- orbital crest or where a small wart was present between its anterior edge and the post-orbital crest. Cranial crest, anterior width: cranial crests are raised bony structures that lie between the eyelids on the top of the head. They are more or less complete from the anterior edge of the eyelids to at least their posterior edges (and often beyond, especially in B. a. americanus). In B. a. americanus they tend to slightly converge anteriorly and flare widely posteriorly; in B. a. hemiophrys they may be more-or-less parallel (sometimes slight- ly bulging or contracting in the middle) but generally diverge anteriorly and slightly or strongly converge posteriorly. The anterior width of the crests was measured across their exterior (lateral) edges at approximately the level of the anterior edge of the eyelids. Some dif- ficulties in precision are inherent, as the crests, particularly in B. a. hemiophrys, merge with the bony crest of the nose, and the latter begins to spread just anterior to the reference point used. In B. a. ameri- canus the nose crest is less prominent and the anterior portion of the crest is more distinct. Despite the potential for incon- sistencies in measurement, the crest dif- ferences are sufficient to make this mea- surement one of the most apparent distinctions between taxa. 10 10. ih 12. ee 14. 15-18. Cranial crest, posterior width: taken across the posterior lateral cranial crest edges, just anterior to their junction with the post-orbital crest or (in many B. a. hemiophrys) just anterior to the post- orbital ‘‘wing’’, a short right-angled spur more-or-less in the plane of the post- orbital crest. In B. a. hemiophrys the crests often do not extend past the post-orbitals and therefore a more posterior measurement, which would have accentuated the crest divergence in B. a. americanus, was not possible if consistency in measuring the two taxa was to be maintained. Post-orbital crest: three arbitrary categories were based on the relative prominence of the bony crest that extends behind the eyelids, from the lateral edges of the cranial crests, at or near their posterior ends, to the tympanum. 0) - post-orbital crest absent or nearly so (not including in this evaluation the short spur on the lateral edge of the cranial crest present in some B. a. hemiophrys). 1) - an intermediate condition between 0 and 2, where the crests are slightly evident, either low and relatively in- distinct throughout or else broken on one or both sides. 2) - post-orbital crests elevated and generally distinct. Eyelid length: the anteroposterior length of the protruding pad of skin above the eye. Head width: taken across the posterior portion of the head of the level at the tym- pani and between their upper edges. There are bony ridges along the upper edges giving a relatively solid structure to measure against. Tympanum diameter: the greatest diameter of the eardrum, usually taken on a vertical plane. Parotoid gland, right and left sides: the parotoid gland lies roughly over the shoulder of the toad and is generally an elongate, somewhat bean-shaped or oval gland. Generally, the margins are distinct in B. a. americanus but in B. a. hemio- Dhrys they may be indistinct, especially at the lateral and posterior edges. Adjoin- 15, 16. 18. 19. BO;)21.. 2223: ing warts sometimes merge with the borders of the parotoid gland, particular- ly posteriorly, and make definition of gland boundaries difficult. The criterion applied for wart inclusion or exclusion in measurement of the parotoid was to omit them whenever possible, that is, whenever they had any appearance of being a distinct wart, whether or not they did ac- tually touch the parotoid gland. Often the left and right parotoids were unequal in size. Separate measurements were taken for left and right sides. Length: the longest axis of this gland. Width: taken at right angles to length, at the widest point. Parotoid separation: taken between the inside edges of the two parotoid glands at the point where the left and right glands approach nearest to each other. Spot length, right and left sides: a measure of the size of markings on an in- dividual toad and the various relations of this to wart size and number. The spot that is usually present on the back at the level of the posterior edges of the parotoid glands and adjacent to the midline on each side was chosen. In rare cases where this was absent the nearest spot to its posi- tion was used. There is considerable in- dividual variation in spot size and shape in both species and in intermediates. The spot is most frequently round or oval but irregular shapes also occur and there is a variable degree of fusion of adjacent spots. Any complete separation, no mat- ter how narrow, was used to define the boundary of this from adjacent spots, but generally constrictions of varying degrees were ignored. In some cases, the spot would be quite regular but in others it would extend far beyond the usual posi- tion. Occasionally a spot was continuous far posteriorly on the back and sometimes anteriorly to the post-orbitals. Only rarely did spots cross the midline. Spots were usually aligned anteroposteriorly, parallel to the midline. In all cases this measure- ment was taken as the longest axis regardless of spot orientation. Spot width: right and left sides: the widest width, that is, the widest portion of the shorter of the two axes, subject to similar 11 23. 24, 25. 26; (21. 29,80: ah, variation to that described in the above section. Tibia: the length of the bone formed by the fusion of the right tibia and fibula in toads, measured from knee to heel with the leg flexed, as it is in most preserved specimens. The measurement taken will be slightly larger on preserved material than on the cleaned bone itself because of inclusion of adherent skin. Minor ad- ditional error may be introduced by dif- ficulties in completely flexing the leg in preserved material and inadvertently in- cluding some of the adjoining articulation of other leg bones. With care this can be largely avoided. The right tibia was measured in all but a few specimens in which it was obviously deformed, broken, or missing. In such cases the left was used instead. Wart width, right and left sides: the width, or largest diameter, of the largest wart in the above spot. Generally warts are more or less circular, but in the rare instances of elongate warts the width was taken across the wart in the same axis as the width of the spot. Number of warts per spot, right and left sides: the number of warts present in each spot defined above. Difficulties due to ap- parent merging or fusing of adjoining warts were resolved by counting as dis- tinct warts each one defined by a separate peak or each portion separated by a distinct crease. Lack of complete separa- tion of warts is most common in B. a. americanus. Diameter of the largest wart of the tibia, right and left tibia: the longest dimension of the largest wart on the upper surface of the tibia. If two obvious peaks were present the wart was regarded as fused from two warts and the measurement taken across the wart so as to measure only the larger one. However, elongated warts which had no obvious peaks were considered one wart and the measurement taken on the long axis. Tarsus: measured from a point on an im- aginary line across the posterior (prox- imal) edges of the metatarsal tubercles (see below) to the ‘‘heel’’—the junction of the tarsus and tibia. There is a certain amount of measurement variation in- troduced because of difficulties in flexing the tibia clear of the end of the tarsus in some preserved specimens. Spade width: the projection used for dig- ging, largest of the two metatarsal tuber- cles on the outer edge of the tarsus, measured across its base from the inner (medial) corner to the outer (lateral) edge of the peak. This measurement is an at- tempt to express the effective digging edge width. It may not, as taken, express the full width of spade but is chosen because of the clarity of its reference points. Tubercle width: the smaller, medial, metatarsal tubercle on the tarsus mea- sured across its base. Venter: a scoring system based on Blair (1943) for the relative amount of the undersurface covered by dark markings (see Figure 5). 0) - no markings on the ventral surface (exclusive of the throat) a2 33: 34. 1) - spot in the pectoral region only 2) - spots covering the chest region only 3) - spotting over the anterior 1/3 of the venter 4) - spotting over anterior 2/3 of venter 5) - spotting over the entire venter 6) - reticulated pattern over the venter This system, unfortunately, excludes much additional variation in ventral markings. The spotting varies in intensity from dark to faint. However, because the variation due to fading in preservative could not be distinguished from natural variation no attempt was made to quan- tify intensity for geographic analysis. There are also differences in the size of the irregular ventral spots and in their relative density on the undersurface that were also ignored in this analysis. Some additional characters, found useful by previous authors in comparing variation within or between these taxa, such as coloration (Gaige 1932; Ashton, Guttman and Buckley 1973; Conant 1975, to differentiate B. a. copei), width of mid- dorsal stripe, height of boss, ulna length (Porter 1968, to differentiate B. h. baxteri), parotoid distinctness (Breckenridge 1944; Henrich 1968), appearance of warts, snout length, head length, pattern of dorsal blotches (Henrich 1968), snout length and inter-orbital distance (Underhill 1961) 12 were not included in the present analysis because of difficulties in objectively measuring or evaluating them, or because the variation they describe is accounted for by a different approach here. Of the omissions, coloration is the most serious because obvious differences are apparent between the two taxa. However, intermediates show such a complete gradation that an attempt at objective categories was abandoned. The character is useless with preserved material because of fading. Call Analysis During the 1969 and 1970 field seasons, tape re- cordings of breeding calls of individual male toads were made on a Nagra III tape recorder. A hand- held omni-directional microphone (Electro-Voice model 635) proved adequate because in virtually all cases toads could be approached and the microphone held within a few centimetres of the calling individual, effectively screening out adja- cent members of the chorus. Occasionally adja- cent toads were transferred to another portion of the breeding pond when they were too close to the individual being recorded. An effort was made to obtain at least four com- plete, clear calls from each individual recorded. Most individuals were recorded at the ponds along the Trans-Canada Highway transect. In 1969 ef- forts were concentrated along the latter transect from about 3 miles [4.8 km] west to about 31 miles [49.9 km] west of the junction of Highway 1 and Highway 11 along Highway 1. Only the single transect was monitored for intermediate calls because of the time required to obtain individual calls. The assumption was made that it was poten- tially more informative to have many individuals along one complete transect than a few from each of several transects. In 1970 call sampling was augmented on this transect and localities west in the range of B. a. hemiophrys to The Pas, Manitoba, and east in the range of B. a. americanus to Oxdrift, Ontario, were also sampled. The body temperature of the recorded toad, as well as adjacent water and air temperatures, was taken immediately after recording was completed. Initially, body temperatures were taken by insert- ing the probe into the mouth and into the vent but it soon became apparent that the difference be- tween temperatures taken by the two methods was essentially negligible and the first method was abandoned. Temperatures were taken with a Yellow Springs single channel tele-thermometer (0°-50°C) equipped with a thermistor probe of small flexible vinyl (YSI Probe 402). The latter was inserted through the vent into the body and the temperature read as insertion was completed. Every effort was made to insert and read the temperature as soon as the animal was captured in order to minimize the effect of struggling while being held (and thus transferring human body heat). In general this proved practicable. Individual toads were tagged (by a numbered tag tied to the right hind leg) after recording and the tag number read with temperature, date and locality directly on to the taped call for that in- dividual. The number read was subsequently used- for the sub-number in the series in which that toad was preserved, thus providing a means of identi- fying the individual preserved toad with its call and allowing comparisons between morphological parameters and call parameters. Tapes were analysed in the laboratory on a Sona-Graph 6061A 85-8000 cps spectrum analyser manufactured by Kay Electronics, Pine Brook, New Jersey. The Sona-Graph produces audiospec- trograms of a selected segment of a call. The seg- ment used in the present analysis was chosen from the approximate mid-point of one uninterrupted call for each individual, after test samples of several calls each from selected individuals in- dicated no difference between calls for the variables considered. Zweifel (1968:271) also noted, in a study of B. a. americanus and B. w. fowleri, the same lack of difference between calls given by the same individual. The narrow band (45 cps filter) gave sharp enough resolution of pulse rate for most B. a. americanus and for B. a. hemiophrys at low temperatures, but wide band displays were necessary to distinguish pulses in B. a. hemiophrys recorded at warm tempera- tures. The number of pulses in a one second in- terval were counted directly from the audiospec- trogram produced. The mid-point of the dominant frequency band was measured by means of a sec- tion to the nearest 100 cps. This is produced by the Sona-Graph as a graphic representation of the relative amount of energy found at different fre- quency levels at a selected point in a recording. These sections were generally taken about half- way through a call. All calls recorded by a given individual were timed by a stopwatch and the mean duration of the call over the sample determined. 13 Computer Analysis of Data All measurements were transferred to computer punch-cards, with two cards required for the data from each specimen. Analysis was run on the IBM system 370 at the University of Toronto Com- puting Centre. The discriminant analysis program for two groups used for comparison of B. a. americanus and B. a. hemiophrys was programmed at the University of Toronto by D.M. Power, November 1967, revised December 1970 and December 1972. The procedure for this program had been taken from Anderson (1966). Discriminant function analysis requires the use of two groups chosen a priori. This was met by excluding all samples from southeastern Manitoba, well to either side of the area of peak interaction noted in field sampling. The reference samples were all collections east and west of this area, chosen therefore on geographical origin without regard to morphology. The weights produced by the analysis were then used to pro- duce scores for the unknowns from the excluded area. Scored characters were omitted from the discriminant analysis and males and females were run separately on the remaining 30 characters (all representing measured values except for two counts—the number of warts per spot on the left and right sides). Because the computer would ac- cept only a maximum of 500 specimens in each reference sample with this many characters used, and over 800 were available in each of the reference samples (819 B. a. hemiophrys and 825 B. a. americanus) for males, the program was set up to run every other specimen in one run for each taxon and then re-run a second time with the alter- nate specimens. Each run compared approxi- mately one-half the sample of one taxa with one- half the sample of the other. Group I included 412 B. a. americanus and 409 B. a. hemiophrys, Group II consisted of 413 B. a. americanus and 410 B. a. hemiophrys. The two sets of discriminant weights were then averaged and the resultant weight for each character was applied to the in- dividual scores for all B. a. americanus and B. a. hemiophrys in the reference samples and to 2460 male Bufo from southeastern Manitoba. Means, variances, standard deviations, and standard er- rors were calculated for each collection. In addi- tion, histograms of the individual scores for each collection were generated. The same program was run for females. The smaller female samples (59 B. a. americanus and 118 B. a. hemiophrys) were included on one com- puter run and the resulting weights applied to 225 females from southeastern Manitoba. A sample of immature individuals proved too small to yield significant results. Although the taxa were distinctive when compared, the number of characters exceeded the number of specimens in one reference sample. Programs for discriminant analysis comparisons of B. a. hemiophrys vs. B. boreas were also run in the University of Toronto program. Additional discriminant analyses involving morphological comparisons of B. a. hemiophrys vs. B. cognatus and the call comparison of B. a. americanus vs. B. a. hemiophrys were run on the National Museum of Natural Sciences, Ichthyology Section, Hewlett-Packard 9830 mini-computer ‘“ICHTHOS” using Program 29: Hotellings T? and discriminant function for two groups from Davis (1971: 284-291) as programmed by D. E. McAllister. ; Analysis of call variables and a discriminant function involving these were also computed on *“ICHTHOS” (McAllister, Murphy and Morrison 1978) as well as the correlation of the discriminant function thus obtained for call and the one previously generated on morphological characters for B. a. americanus, B. a. hemiophrys, and inter- mediate populations. Hybridization Experiments During 1969 calling males and gravid females were collected for mating experiments from choruses along the Trans-Canada Highway transect through the interbreeding zone. The localities used were: (1) 31-29 miles [49.9-46.7 km] W. on Highway 1 or junction of Highways 1 and 11: B. a. hemiophrys (2) 17 miles [27.4 km] W. on Highway 1 of junc- tion of Highways 1 and 11: intermediate population. (3) 3 miles [4.8 km] W. on Highway 1 of junc- tion of Highways 1 and 11: B. a. americanus A total of nine crosses with three replicates of each (27 sets) were attempted. female male americanus xX americanus americanus x Intermediate americanus x hemiophrys intermediate x Intermediate intermediate x americanus 14 intermediate x hemiophrys hemiophrys x hemiophrys hemiophrys x intermediate hemiophrys xX americanus Males and females were collected the same night from all three localities and paired in selected com- binations in four- or 10-litre jars in pond water from the locality where the female had been col- lected. Mating and egg laying usually took place overnight but occasionally was delayed until the next day. If it did not occur within 48 hours the cross was repeated with another pair. Two hundred eggs were counted out from the batch laid by each female of a successful cross, and divided into four lots of 50 eggs each. One set of 50 eggs from each cross was raised in the laboratory, one of each other lot was placed in a container in each of the three sites from which the adults had been collected. Laboratory crosses were raised in dechlorinated water in shallow plastic pans, field crosses were initially placed in circular petri dishes, each with a piece of their lid removed and replaced by plastic screening to permit water circulation. After hatching, the tadpoles were released into open plastic rearing cages made from plastic screening with a framework of plastic rods. Each unit was approximately 120 x 120 x 30cm and each included four 30 x 30 x 30cm com- partments screened on the sides and bottom but open at the top. They were placed in the water at the edge of the ponds and submerged to approx- imately half their height. Receding pond levels due to evaporation during the period from hatching to transformation of the tadpoles and level rises after occasional heavy rains necessitated moving the units to compensate for these fluctuations in water levels. Initially, the eggs were checked at daily intervals, but later in tadpole development containers were checked at two-day intervals, until the last tadpole had metamorphosed. Tadpoles were fed Tetramin, a commercial ‘‘fish food’’ preparation that does not unduly contaminate water. The water in laboratory pans was changed whenever it appeared fouled but this was not necessary in field cages. Dead eggs and tadpoles were removed and noted at each inspection. Part way through the experimental period a disaster oc- curred at the ‘‘americanus’’ site when all rearing cages were removed from the water and placed neatly at the roadside by person or persons unknown. The majority of tadpoles then desic- cated due to exposure to air before the next regular inspection, thus terminating these crosses. Results 1. Field Surveys Field work in eastern Manitoba in the breeding seasons of 1968, 1969, and 1970 revealed no strong differences in breeding times, either in the initia- tion, peaks, or termination of calling between B. a. americanus, B. a. hemiophrys or intermediate populations. B. a. hemiophrys localities tended to be in more open sites, and the air and water temperatures dropped there more rapidly on cool evenings, resulting in B. a. hemiophrys choruses tapering off or stopping earlier than B. a. ameri- canus choruses. The general pattern evident in the 1968-70 field seasons in eastern Manitoba was an _ initial emergence from hibernation and arrival of an occasional individual at the breeding site as early as 19 April but as late as early May. The earliest calling was 24 April in 1969, with the earliest amplexed pairs and eggs 6 May the same year. The first good choruses were heard 13 May 1968, 14 May 1969, and 17 May 1970, indicating that an early or late spring does not alter the time of the initial peak of breeding by more than a few days. It was common in all years to find good choruses on warm nights from mid-May through to early June and for calling to gradually taper off with a few persistent individuals calling as late as the first week in July. A few small cheruses were found in middle and late June in 1968 but the bulk of the series from breeding ponds were taken earlier (see list of specimens examined in Appen- dix I for dates of strong choruses indicated by size of collections). A similar pattern emerges when the collection dates of breeding B. a. hemiophrys from throughout its range are examined. One large collection taken later in the year (60 specimens from the Spruce Woods Forest Reserve 20-21 July 1960) is a post-breeding sample from toads along a sandy road. All large breeding samples are from the latter half of May or early June. It is notable that there is a much greater varia- tion in the dates for B. a. americanus breeding series, especially the series (NMC 6916) obtained adjacent to Rondeau Park in southern Ontario on 17 April 1963. In the milder climate of south- western Ontario toads apparently may reach full chorus as much as a month earlier than is possi- ble in eastern Manitoba or the Prairie Provinces 15 in general. In eastern Manitoba, there was no con- sistent pattern during this study of either B. a. americanus or B. a. hemiophrys beginning call- ing appreciably before the other, or continuing later. The latest chorus sampled was of B. a. hemiophrys-like individuals (Grand Beach Provin- cial Park) 21 June 1968 but there were B. a. americanus choruses at N.W. Angle Forest Reserve a few days earlier, 16-17 June 1968. In both cases the proximity of the breeding localities to large lakes may have retarded the start of the breeding and prolonged it in comparison to other localities. Bufo started to metamorphose in early July in eastern Manitoba. The earliest noted was 2 July 1969. This coincided with the cessation of calling from the last few persistent males and represented roughly a month and a half for development after the initial peak breeding. The latest tadpoles were noted 5 August 1968, but most had transformed by the last week in July, representing a two-three week span of peak transformation that roughly corresponded with the span of most active breeding. Breeding ponds for both B. a. americanus and B. a. hemiophrys in eastern Manitoba and else- where are variable. Artificial dugout ponds, road- side ditches, and shallow quiet shores of rivers and lakes are common breeding sites. A fairly wide area of shallow water seems to be preferred. The breeding sites for B. a. hemiophrys, being mainly in aspen parkland and prairie, are in more open surroundings due to the general habitat, and no conclusions can be drawn on any difference in site preference. In general, the same types of breeding sites seem to be chosen by the two taxa. In 1970 a simple mark-recapture program was carried out at the intermediate (contact zone) locality 17 miles [27.4 km] west of the junction of Highways 1 and 11. Fifty-one male toads were marked and released between 17 May and 11 June. An additional 23 males were collected 18 May from this locality and preserved. Forty-nine recap- tures of 28 individual toads were made. A total of 16 females was also taken during the same col- lecting period. Some males remain at the chorus through the peak breeding period although un- marked individuals may be arriving throughout. Four individuals marked and released 18 May were recaptured on each of 22, 23, and 24 May and three others were captured on two of those nights. There was no evidence of an uneven distribution of morphological types between early and later samples. A simple population estimate gives figures of 42, 47 and 64 male toads for 23 May, 24 May, and 5 June, respectively. 2. Analysis of Morphological Variation for Reference Samples a. Discriminant Analysis of Males The two discriminations run on alternate specimens from the combined reference sample gave the weights or character coefficients shown in Table 1. A mean discriminant weight for each character was produced by averaging them (Table 1) and these mean weights used to compute discriminant scores for each individual. The size of the weights reflect in part the relative importance of the characters considered in separating the taxa. The relative effectiveness of the weights is dependent on the number of units an individual character varies over, and the separation between the taxa within this character. A method of ex- pressing the general distance between taxa is the ‘‘Mahalanobis generalized distance’’ computed by finding the separation between the mean discrimi- nant scores of the taxa. The contribution of each character weight to this value can be obtained by multiplying the weight by the difference between Table 1. Discriminant weights given each character in two analyses (Groups I and II) of male reference samples of Bufo a. americanus and B. a. hemiophrys, and the mean weight of the two groups used in computing discriminant scores for individual males, and the weight contribution to the Mahalanobis generalized distance between the combined groups. Weight Group I Snout-vent length + 0.16522861 Nostril separation — 8.1142540 Cranial crest length (right) + 0.76029348 Cranial crest length (left) + 3.6722383 Nostril to parotoid (right) + 0.41164047 Nostril to parotoid (left) — 0.59715325 Cranial crest width (anterior) — 4.7215214 Cranial crest width (posterior) + 8.3745794 Eyelid length — 3.5560398 Head width — 2.0142698 Tympanum diameter — 0.65865707 Parotoid length (right) + 0.073451102 Parotoid length (left) — 0.48348296 Parotoid width (right) + 1.0284834 Parotoid width (left) — 0.89462119 Parotoid separation — 3.0141783 Spot length (right) Spot length (left) — 0.034990408 + 0.055400848 Spot width (right) — 0.37663031 Spot width (left) — 0.76379973 Wart width (right) + 3.37714357 Wart width (left) + 3.5572758 No. warts per spot (right) No. warts per spot (left) Tibia length Diameter of largest wart on tibia (right) Diameter of largest wart — 0.090008795 — 0.098180711 — 1.1040401 + 0.29262280 on tibia (left) + 1.0001144 Tarsus length + 4.0203476 **Spade’’ width — 6.1017218 Inner metatarsal tubercle width + 3.0812159 Weight Mean Mean Group II Weight Contribution + 0.18881679 + 0.177023 + 0.9242126 — 8.586230 — 8.350242 + 2.5555703 + 3.2909012 + 2.025597 + 3.5536756 + 0.50351000 + 2.087874 + 3.6579865 — 0.78529143 — 0.186825 — 0.2177316 + 1.0789080 + 0.838030 + 0.9472763 — 4.5984650 — 4.659993 + 0.2671204 + 8.6095295 + 8.492055 + 21.7769295 — 3.2611437 — 3.408592 — 1.4677809 — 1.8766050 — 1.945437 — 1.3364795 — 0.21684265 — 0.437750 — 0.3360502 + 0.31809098 + 0.195771 + 0.2395683 — 0.37021834 — 0.426351 — 0.4372118 + 0.13527966 + 0.581882 + 0.0483071 + 0.41438520 — 0.240118 — 0.0093115 — 2.6087561 — 2.811467 + 4.1812830 + 0.036637843 — 0.000824 — 0.0008016 + 0.025961559 + 0.040681 — 1.0455871 — 0.18416786 — 0.280399 + 0.0474776 — 0.31844270 — 0.541121 + 0.0950692 + 1.0056877 + 2.388562 + 1.8619682 + 1.7886143 + 2.672945 + 2.1319337 — 0.10848355 — 0.099246 + 0.1581416 — 0.18084419 — 0.139512 + 0.2189280 — 0.79755020 — 0.950795 — 2.1168822 — 0.93891734 + 0.615770 + 0.5780064 + 1.13967094 + 1.198411 + 1.1220402 + 4.3927183 + 4.206533 + 10.3077514 — 6.1168079 — 6.109264 + 1.9970081 + 0.45624775 + 1.768732 + 0.6493347 Mahalonobis generalized distance (mean): 51.3517513 16 the means of each taxa for that character. This gives the weight contribution (to the Mahalanobis generalized distance) and the sum of these con- tributions is equal to the Mahalanobis generalized distance (D.E. McAllister, personal communica- tion). This weight contribution to separation of the taxa is also given in Table 1. Mean values for each character. are given in Table 2. Posterior cranial crest width, tarsus length, parotoid separation, and cranial crest length make relatively large contributions (in descending order) to the separation between taxa. High values for nostril separation, anterior cranial crest width, head width, parotoid separa- tion, eyelid length and width of the outer metatar- sal tubercle (‘‘spade’’) seem typical of B. a. hemiophrys, and high values of posterior cranial crest width, cranial crest length, width of largest wart in a blotch, tarsus length, and width of in- ner metatarsal tubercle seem typical of B. a. americanus. When average discriminant weights were ap- plied to the entire male reference sample of both species (1644 specimens) to obtain a discriminant value for each specimen, a single specimen, a B.a. hemiophrys from Delta, Manitoba, was misclassified, due to an error in measurement which was discovered on re-examination of the specimen and corrected. Recalculation with this measurement corrected gave a discriminant score of — 10.2223, well within other B. a. hemiophrys scores. Table 2. Mean values for each character measured in reference samples of male Bufo a. americanus and B. a. hemiophrys. Bufo a. americanus Bufo a. hemiophrys Group I Group IT Group I Group II Snout-vent length 62.159164 62.004028 56.797302 56.924149 Nostril separation 3.9171495 3.9309130 4.2321920 4.2279654 Cranial crest length (right) 12.622904 12.582693 10.848487 10.848342 Cranial crest length (left) 12.634803 12.616809 10.868276 10.879306 Nostril to parotoid (right) 12.759092 12.718702 11.580885 11.566053 Nostril to parotoid (le*t) 12.764392 12.719232 11.629278 11.593626 Cranial crest width (anterior) 4.9079227 4.9357510 4.9805956 4.9777222 Cranial crest width (posterior) 6.7011232 6.6863594 4.1363468 4.1223583 Eyelid length 7.5224867 7.4919243 7.0796156 F0735712 Head width 17.055084 17.057892 16.364105 16.374908 Tympanum diameter 4.9554949 4.9359989 4.1847582 4.1713829 Parotoid length (right) 11.721546 11.619533 10.434548 10.459098 Parotoid length (left) 11.711373 11.650821 10.654340 10.659307 Parotoid width (right) 6.4105911 6.3711004 6.2896471 6.3260069 Parotoid width (left) 6.3710337 6.3008833 6.2554150 6.338944 Parotoid separation 7.7693291 7.8577881 9.3114061 9.2901602 Spot length (right) 8.2746706 8.1989489 9.3226452 9.0972290 Spot length (left) 8.0421515 7.8604441 9.1226511 9.0211334 Spot width (right) 4.3023443 4.2669897 4.4395208 4.4684544 Spot width (left) 4.3076839 4.2541599 4.4869576 4.4262648 Wart width (right) 2.7535648 2.7376938 1.9527435 1.9794445 Wart width (left) 2.7809954 2.7781324 1.9698620 1.9940720 No. warts per spot (right) 2.7038832 2.9176750 4.4498777 4.3585358 No. warts per spot (left) 2.6941738 2.6973362 4.2933979 4.2365847 Tibia length 23.107529 23.026627 20.834091 20.847198 Largest wart on tibia (right) 3.0771122 3.0401163 2.1119099 2.1279736 Largest wart on tibia (left) 3.1288061 3.074981 2.1693630 2.1618786 Tarsus length 15.663939 15.685769 13.216190 13.232688 ‘**Spade’”’ width 4.7040462 4.6698914 5.0246153 5.0030861 Inner metatarsal tubercle width 2.4746885 2.4432659 2.0818386 2.1018782 Mean discriminant score (centroid): 24.037781 24.683716 — 29.128174 — 24.89318 Mahalanobis generalized distance: between Group I centroids: 53.165955 between Group II centroids: 49.578034 A? B10 O-+ ELE e100 +9870 L70'0 + CST'0 800°0 + €LE°0 CVFEGCS 00+0'S 6001 0OF0'T 6s? 6.1c— 6 Nd 070°0 + 86€°0 810°0 + 8970 87Z0'0 + 8¢1°0 O10'0+ PSE°0 She Vals 60497 9°0+60 v0+8°0 Lec. TTC ve WS 8200 + ELE0 070°0 + 1870 870'0 + 9ST 0 600°0 + cSt°0 OE +8'SH LiFe v0+60 v0+8°0 O's! Bak 60 WY? - 910°0 + 86£°0 670'0 + 7870 0£0°0 + P9T'0 800°0 + 6v£'0 OE +B'SS $'0+0°S L:0+8°0 0'°0+0'T S8l LOCH 8 We $cO'0+ PLEO TZ0°0 + OS7'0 8£0°0 + €9T°0 010'0 + S9€°0 GLFVES cTOFO'S LOSeOr C001 ole ale— 07 WC 170'0 + CSE'0 ¥70'0 + SP7'0 c£0'0 + T9T'0 600°0 + 79€°0 DYFELS 0'0+0'S V0 El €:0+6°0 COC ICE vl WI 120°0 + 69£°0 0€0°0+ L¥7'0 T€0°0 + 7910 0100+ £9€°0 tat 0 LY, SO+6P7 EO 10 as (a0) S8c— oO9C= DL S6 cE0'0 + S8E°0 170'0 + v$7'0 0€0'0 + 8ST'0 O10'0 + LLE'0 VCS CO 9°0+8°7 $'0+8°0 $°0+9°0 OTe 0b VC S8 970'0 + O8€'0 €c0'0 + 8770 cv0'0 + P9T'0 T1100 + €L€'0 Leste’ SO0+6P? 9°0+8°0 5.0+9°0 a Slay Sa Ga Le eye v70'0 + ILE°0 L1I00+%S7'0 T€0°0 + €ST'0 0100 + 9LE'0 CEO CS 90+F8'P LO el c0+6°0 CSc G9C— 9¢ S9 1€0°0 + 88£°0 910°0 + 8970 ve0'0 + OST 0 T10°0 + 19€°0 LU Si cl Se 0';0+0°S ¢'0+S°0 vO+T0 LENG S6G= CG SS 070'0 + 68€°0 LI0';0 + S€7'0 v70'0 + TST 0 810'0 + 19€°0 GE EOS S067 oO+S5-0 €'0+60 SL om S6C— (@E SP 9720°0 + I8€°0 €70'0 + 0E€7 0 ¥70'0 + P9T'0 LLOIO ELE 0 vers 6P 60+F+6P7 ¢0+¢0 vO+L0 IGS Cue 88 SE 970°0 + 00F'0 c7O0'0 + 8S7'0 L700 + 8ST'0 800°0 + €S€°0 LCC 9S 80+9'P D0 2.0 vO0+70 lekc Sa Ge OV Sc $c0'0 + 9LE'0 170°0 + 1S7'0 €70'0 + 8S1°0 vI0'0 + S9E°0 OCC EES. 0'0+0°S 9°0+6°0 c0+8°0 Pics Ese 81 ST CEO U+TLE-D 170°0 + OVZ'0 6700 + I9T'0 600°0 + PLt'0 ty 1 $9 ORCS, oO £0 v'0+6°0 SVC ee 89 V8 ¥20'0 + CO'0 9S0°0 + 7970 ¥Z0'0 + 6ST 0 010°'0 + 99€°0 Adres) VO-rCS OF 1-0 $'0+9°0 Dao Gk i ST Soa 07 VL c70'0 + 96€°0 $c0'°0 + L970 870'0 + 7910 1100+ I8€°0 OE OL 0'0+0°S 9°0+6°0 S0+9'°0 Sie -FoOEe= OF v9 970'0 + 8S€°0 L700 + IL7'0 ¥20'0 + T9T'0 c10'0 + I9€°0 CV+8 ps 0°0+0°S 04 9°0 S02 0 SEGe V ce = EC vs 170°0 + LLE'0 610°0+ Lv7'0 070°0 + 8910 800°0 + 19€°0 9TFOLS CO +05 $'0+S°0 y0+C0 PALES S2Lo- gs Vv 670'0+ PLt'0 €50°0 + 6970 970'0 + SLT'0 1100 + S9€°0 Cte FS 0°0+0°S 20+ & 1 p0+80 Oot e eee ot ve 1Z0'0 + 9E"0 970°0 + 9¢7'0 620'0 + 6910 c10 0+ 89€°0 ye+sS'09 €0+0°¢ S0+v0 ¢'0+9°0 Grit LEG 8P VC 110°0 + L6E°0 070°0 + 097'0 0£0°0+ LST'0 010°0 + 69€°0 Oc FE 09 0°0+0°S 90+ 6.0 COTE 0 Bela Ortec= 8 Vi ¥20'0 + 78E°0 €70'0 + 9S7'0 1€0°0 + 19T'0 ELO0 + TLE 0 Pees O ES 0';0+0°S 02 8:0 SOL 0 LseCe CVC OF Al sduydoiuay snuvdisaiup ofng SFX S#X SX SORIX SFX S +x SFX Sx 7s Zz dZIS Joquinu one one ouel one YyIsuds] 3109S 3109S 3109S oydwies AN]e90 7 YIsus]/YyIpPIM SUIPIM yi3ud] yis3ud] (AS) 1U3A 3109S [e1qio (95) 189.19 uonouny snsiej/apeds peoy/oo AS/10dS AS/BIQLL -jnous 13}U3 A, -1SOd yetuelg JUBUTUILIOSIG 18 ‘(SUUUINJOS IdYyIO []e) (S) UOTJVIADP plepuR}s 9UO F JO (A10DS UOTJOUN{ JUBUTWILISSIP) (7S) SOUBLIVA JY] YIM UDAIS SI o[dures Yoea 1Oj (X) UBS OU], “JAS /AaJIDADYI [DIIBOJOYAIOPF JIpuUN UOTIDIS SPOYIIJA PUL STRLIDIVI oY} Ul paqiiosap puke ¢ pUe pf SOINSI Ul UMOYS 318 pasn SIdJOVIVYD “UOTDIJ[OO JO sawp dy] pue SJaqUINU dNSoOyeIeO YUM | XIpuaddy ut paisty oie pue Z ainsi ul paddeu ose soryeso] sy ‘(N) B100S BPAON pue (gq) YOIMSUNIG MAN ‘(q) PURIST preMpA doulIg ‘(H) daqGang% ‘(O) oe} “uO “(NIN) BlOsauUT, WOoIJ (SoTeW ¢7Q) Sardues gy] WIJ SNUDIIUWaUID ‘D “Gg Jo pue (QC) B1IOXeq YWON pue ‘() eqoiueRpy ‘(S) UBMaYyDIeYsEeS ‘(V) BLOqIV “(AM\) SUIWIOAM WO (SoTeUI 61) SaTduIes pz Woy Sduydoluay snuvdiiaup Ofng JO SISkyeue UI pasn speo} afeur JO (AIITedO] Aq pajood) saydures aduaJojal JO UOSTIVAUIOD IO} SONBI Pa}daJaS INO} PUB JUdSUWIIINSedU JUO ‘sa1ODS ddIY) ‘UOTIOUNJ IURUTWILIOSIG “¢ aIqUyL c£0°0 + LOE 070°0 + €0F'0 pe0'0+OIT 0 v10'0 + 89€°0 Gtrra 6047S © 0+ 6 00+0E€ SC. Scctr OV Nd 610°0 + 867°0 ¢70°0 + 00b'0 6£0°0 + SET 0 110'0+ 78€°0 OC+C ED sO0FTCS CO+0'C OOF O' ee. 9g Ss NI I70'0+ PIE 0 1700 + 86£°0 970°0 + SOTO L10';0+SLE'0 Sere LS GOFGS COFOC 00F0E O8t BCC + €s dl 1Z0°0 + ITE°0 1700 + POr'0 €£0°0 + 6110 c10'0 + S8€°0 ova vo S096 0'0+0°7C O'OF0OE £32 P6tr 6P dl $10°0 + L670 ¥70'0 + 88£°0 ell O0+9I70 v10'0+ 8ZLE'0 PCr 81h 9-0L-5 0'0+0°7 00+0'E b6p O8c+ CC OS L10°0 + $670 $700 + 6LE0 €v0'0 + OFT 0 vI0' 0+ SLE0 OP ts 99 LLtL Vy LOFOC 00+0€ OCh 8 H+ bs Or 610°0+FI1E0 970'0 + Crh'0 $S0'0 + 9ET 0 8100+ vSE'0 LE ces c€0+09 0'0+0°C 00+0°€ Lk L6G 8f Of v10°0 + LOEO 6c0°0 + Ttr'0 890°0 + L¥I'0 600°0 + IS€°0 6€+6 79 €0+6S 0'0--0'¢ 00+0'¢ cco Viper ce O07 670°0 + L970 610°0 + T8E°0 1S0°0+ SPIO €10'0+ L9E'0 PE CSS vO+6'S v0+8'I 00+F0'E Oe. fet vl OI 970°0 + 887°0 $10°0 + COP £700 + £60°0 1100 + 78€°0 ELL Lo eat Mia 00+0°7 0'0+0'€ O'S Ser Ol O8 1700 + €87°0 170°0 + 8070 670'0 + 680°0 c10'0 + S8€'0 697159 OTFCC vO0+6'T 00-0 E Ly OLEs LV OL ¥70'0 + 687°0 7700+ £6€°0 €r0'0+ PET 0 110°0 + O8€°0 Ceres 09 CLs? v0+F6'l OOFOE 6c. O9¢+ 68 O9 $700 + 667'0 070°0 + 06£°0 $v0'0 + 8710 B10'O= £9t 0 EPes vg 20+ 0 20-611 00+F0'°€ PSE £9¢C+ vol OS €70°0 + SOE0 870'0 + S6E°0 Sv0'0 + 8710 vl0'0+7LE°0 SVTL YY 20 1S £ 0-56 ft 0 O+0't bss Sver 9S Ot 910°0 + 867°0 1£0°0 + 6££°0 $r0'0+ FIT 0 Sv0'0 + TSE0 OPFP 8S O1+8'P 907751 00+0¢ O8& Felt vs Of L70'0 + 667°0 $70'0 + OLE0 9S0°0 + PST'0 710'0 + TLE'0 8'7+0°CS COFLS 0'°0+0°7C 0°0+0°€ ose £9b+ 09 O7¢ 070°0 + SOE*0 L700 + 6LE°0 O€TOFSITO €10'0*8LE'0 [S$ +6-09 SO0+*8'S p08 £0F67C Sor .Stcr 81 Ol $10°0 + 90£°0 970'0 + O8€°0 0v0'0 + 801°0 c10'0 + £90 8°7 + 9°09 O1+6'7 CcOFOTC 00F0'E' Ove S List O€ NWI SNUDIIAAUD SnNUDIIIaWD Ofng SFX SFX SFX SFX i x SX SFX S + xX 2s x oZIS loquinu one one one one yIsud] 3109S 3109S 3109S a]duies Al[e00 FT yisus,/yIpIM SYIPIM yisud| yIsud] (AS) 1U9A 9109S [e1gqio (99) IsoI5 uonsunj snsiei/apeds peoy/do AS/10dS AS/PBIQIL -Inousg JO1UI A -1SOdg yetuelg JUBUIWULIOSIG] 19 The sample size for each population sample of males is given in Table 3. Histograms for the pooled male reference samples are given in Figure 9. The population mean for the discriminant score, ranging from — 18.7 at Delta, Manitoba, to — 30.8 at Qu’Appelle Valley in Saskatchewan for B. a. hemiophrys and + 13.3 at Oxdrift, Ontario, to + 33.8 at Rondeau Provincial Park, Ontario, for B. a. americanus (see Table 3). The highest values for B. a. hemiophrys are in the Aspen Parkland and Boreal Forest of Manitoba and adjacent North Dakota with slightly lower values across middle Saskatchewan and Alberta and northern Alberta, largely in the Boreal Forest. The mean of the disjunct Wyoming population falls with these. Southern Saskatchewan and Alberta populations have the lowest values from shortgrass prairie and Aspen Parkland habitats. The lowest values for B. a. americanus are from northern Minnesota (Hibbing), northwestern Ontario (Ox- drift) and the southern end of James Bay (Whitetop Creek). The variance, as an indication of the dispersion of discriminant values within the population samples, is given in Table 3. Variance of B. a. hemiophrys populations were generally lower than for B. a. americanus, although higher variances are seen in B. a. hemiophrys populations in areas adjacent to eastern Manitoba. Histograms of discriminant scores for each population sample, or pooled sample, are given in Figures 10 to 15. In general, they confirm that the values are roughly normally distributed within each population and that individual values are well away from the mid-point value of the discrimina- tion axis. The closest approach to the mid-point is made by an individual B. a. hemiophrys from Perryvale, Alberta (Figure 10), and an B. a. americanus from Oxdrift, Ontario (Figure 14). b. Univariate and Ratio Analysis of Males In addition to the discriminant function analysis, the geographic variation in three scored charac- ters: cranial crest, post-orbital crest and venter (Table 3), all excluded from the discriminant analysis, was examined. Snout-vent length and ratios of tibia/snout-vent, spot length/snout-vent, cranial crest posterior width/head width and outer metatarsal tubercle width/tarsus length (Table 3), were also examined for patterns of variation. These latter measurements had been used in- dividually in the discriminant analysis. 20 120 110 100 90 80 : = = AMERICANUS LU = -20 =10 0 10 20 30 40 50 60 Figure 9. Histograms of individual discriminant scores for male reference samples: 819 B. a. hemiophrys (open) and 825 B. a. americanus (hatched). Vertical scale represents number of specimens, horizontal scale the discriminant score. Figure 10. Histograms of individual discriminant function scores for eight Alberta reference collections of male B. a. hemiophrys. The two vertical lines indicate the mid-point of the discriminant analysis between B. a. americanus and B. a. hemiophrys. M in- dicates the position of the mean for the sample and D is the position of one standard deviation on either side of the mean (the unequal distances apparent in some histograms are due to rounding values to fit the whole number scale units used). For each sample the locality designation is given, and to the right of it the catalogue number (National Museums of Canada unless other- wise indicated) followed by a dash and the number of specimens in the sample. ALBERTA Mikkwa River 14045-8 0.5 mi. E. on hwy 16 8502-8 Stoney Plain 8505-47 6mi. SW. on hwy 2 Perryvale 8527-36 2.4 mi.N. on hwy 44 Westlock 8521- 48 3.3 mi. SW. St.Paul 8547- 27 10325-3 10332-12 10333-9 10755=12 3.2) Mic N, SC Gimi'E. jet. 4thst.NeE.& Northmont Dr. Calgary 7115 - 14 Strathmore mins 6021-11 0.5-2.2 mi.NW.on hwy 1 6022-31 Brooks (turnoff) 6023-26 The horizontal scale gives the discriminant score values, the vertical scale the number of individuals. 21 SASKATCHEWAN (1) 4049 -4 4052 -1 4060 -1 4068 -5 Prince Albert 4072 -7 6.5 mi. E. on hwy 14 Reward (turnoff) 6007 -40 8.2 &14.5mi.E.on hwy 5 6038 -44 Sutherland 6040 -44 Qu’Appelle Valley 5263 <1 18.0 mi. W. & 15.2 mi. N. 5265 -10 Belle Plain 5267 -13 NW. end Buffalo Pound Lake 4007 -27 2.3 mi.E. on hwy 1 Sintulata (turnoff) 8496 -74 Leach Lake at hwy 9 14.3 mi. S. Yorkton 4033 -22 Figure 11. Histograms of individual discriminant function scores for seven Saskatchewan reference collections of male B. a. hemiophrys. See Figure 10 for explanation of letters, scale, etc. 22 SASKATCHEWAN (2) 3.3 mi. W. on hwy 1 Tompkins ( turnoff) 6014 -36 8.3 mi. E. on hwy 1 Piapot (turnoff) 7111-22 WYOMING, USA Laramie River valley KRP - 30 8610 - 6 Figure 12. Histograms of individual discriminant function scores for two Saskatchewan and one Wyoming reference collections of male B. a. hemiophrys. See Figure 10 for explanation of letters, scale, etc. 23 MANITOBA 6-10 mi. S.on hwy 10 12252 -14 The Pas 8mi. E. on hwy 10 Minitonas (turnoff) 12256 -20 0.5mi. N. of hwy 1on 4563 -3 Oberon (turnoff) 4565 -5 NE. section 4608 - 34 Spruce Woods Forest Reserve Delta Beach NORTH DAKOTA, USA 41 mi.E. of Devils Lake 3338-9 Figure 13. Histograms of individual discriminant function scores for five western and central Manitoba, and one North Dakota, reference collections of male B. a. hemiophrys. See Figure 10 for explanation of letters, scale, etc. 24 ONTARIO 14783 -21 12 mi. NE. Moosonee 14791 -39 Whitetop Creek, Cochrane Dist. 14797-18 Moosonee, Cochrane Dist. Beaver Creek 6 mi.W.on hwy 17 Oxdrift 12238-54 Lake Kenogami 7076-56 10.5 mi. E. on hwy 17 North Bay 7064-104 imi. W. 6958-89 Carlsbad Springs, Carlton Co. 0.5 mi. N. Rondeau Prov. Park 6916-47 Point Pelee Nat. Park 9684-10 MINNESOTA, USA 4 mi. S. on hwy 73 3994-30 Hibbing Figure 14. Histograms of individual discriminant function scores for seven Ontario and one Minnesota reference collections of male B. a. americanus. See Figure 10 for explanation of letters, scale, etc. 25 QUEBEC Lac Attila, James Bay 15851-14 2265-14 2269-12 Zen =F , 2278-1 Mile 134, N 2279-1 Seven Islands 2280-3 2237-24 2238-4 2239-1 2241-1 2242-3 Seven Islands 2258-5 Cap de Rosier 13134-54 Routierville 2429-22 NEW BRUNSWICK 6mi. N. Moncton 8711-53 PRINCE EDWARD ISLAND 0.6 mi.E. 3964-29 Roseneath 39 88-20 NOVA SCOTIA Wolfville 8716-55 Musquodobait Harbour 8729-40 Figure 15. Histograms of individual discriminant function scores for five Quebec, one New Brunswick, one Prince Edward Island and two Nova Scotia reference collections of male B. americanus. See Figure 10 for explanation of letters, scale, etc. 26 Cranial Crest Score: Scores of 0 (boss) and 1 (grooved crest bridged posteriorly) were thought to be characteristic of B. a. hemiophrys. A score of 3 denoted a typical B. a. americanus crest. A score of 2 was assigned to any intermediate con- dition. Mean values for B. a. hemiophrys range from 0.1 to 1.0 and all mean values for B. a. americanus are 3.0. Values for B. a. hemiophrys, from Manitoba and North Dakota samples are consistently high, 1.0 to 0.8. Post-orbital Crest Score: All B. a@. americanus samples examined in this study had mean scores of 2.0 and 1.9 with the exception of 1.8 scores for the Lac Attila and Moosonee samples and 1.5 for Oxdrift. The means for B. a. hemiophrys ranged from 1.5 to 0.1 with high values (1.0 to 0.8) in eastern Manitoba, but considerable variation ex- ists between populations elsewhere in the range, with one central Alberta population (Perryvale) scoring 1.5, and one southern Saskatchewan population (Tompkins) scoring 1.3. Venter Score: This score ranged from 0 (im- maculate venter) to 6 (reticulated venter) and as such covers the variation in B. americanus group toads. Means for B. a. hemiophrys varied from 4.3 to 5.0, with low values in central Manitoba (4.3 at Delta, and 4.6 at Spruce Woods). The highest values were in southern Alberta (5.2 at Brooks and at Sutherland) with northern values in Alberta, Saskatchewan and Manitoba all at 5.0. The range in B. a. americanus was greater, from 4.2 at Rondeau to 6.0 at Sept Iles. Northern B. a. americanus all tended to have high mean scores (above 5.5) and southern had the lowest scores (4.8 and lower), with those intermediate geographically tending toward intermediate scores (5.0-5.2). An anomaly is the Cap-des-Rosiers population at the tip of the Gaspé with a mean of 4.7. Less surpris- ing is the 5.6 score for Roseneath, Prince Edward Island, putting it closer to northern populations in this character. Snout-vent Length: Although there is an obvious difference in means between species (Table 2) for the whole reference sample there is a great deal of variation between population means within each species. For B. a. americanus the range was from 71.8 mm (Rothierville, Quebec) to 53.9 mm (Moosonee, Ontario). B. a. hemiophrys ranged from 70.8 mm (Calgary, Alberta) to 45.8 mm (Delta, Manitoba). For individual males, B. a. PF hemiophrys ranged from 79.1 mm (near Brooks, Alberta) to 41.3 mm (Delta, Manitoba) and B. a. americanus from 88.4 mm (near Rondeau Provin- cial Park, Ontario) to 42.1 mm (Whitetop Creek, near Moosonee, Ontario). Commonly, there is about a 20 mm spread from the largest to the smallest mature male in any given sample, though it ranges from about 5 to 25 mm, with the smaller B. a. hemiophrys generally having less of a spread (often 10-15 mm) within one population sample. However, in the Rondeau B. a. americanus sample the range is 38.4 mm. The largest B. a. hemiophrys sample means (61.2 mm and larger) are in the prairie region of southern Alberta and Saskatchewan. There is no clear-cut geographic or vegetation region arrange- ment of the other groups, though there is a ten- dency for the sample means to be smaller in the east and larger in the west. Within B. a. americanus the smallest means are generally in the north (52.9 to 60.5 mm) but the large value of the sample from Mile 134 [Kilometre 215.7], north of Sept Iles, is out of line at 64.9 mm, and the small value from Moncton, New Brunswick, at 57.3 mm is aligned with the smaller northern means. Tibia/Snout-vent Length: The relative length of the tibia is variable in both B. a. americanus and B. a. hemiophrys. Within B. a. hemiophrys there is little apparent pattern, though southern Manitoba means are among the lowest, and southern Alberta and Saskatchewan are among the highest. In B. a. americanus there is a tendency for northern scores to be low and southern means to be high. Spot Length/Snout-vent Length: The range in B. a. hemiophrys (0.148 to 0.175) is much narrower than that in B. a. americanus (0.090 to 0.216). Within B. a. hemiophrys there is no obvious geographic pattern to the sample means of this ratio from north to south or between vegetation areas. In B. a. americanus, however, the northern populations (Moosonee, Lac Attila, Mile 134 [Kilometre 215.7], Sept Iles, Routhierville and Cap-des-Rosiers) have generally higher values (be- tween 0.136 and 0.216) which encompass the B. a. hemiophrys means, while more southern populations (Oxdrift, Hibbing, Rondeau, Monc- ton, Roseneath and Musquodoboit) have low values (0.119-0.090). Kenogami, North Bay and Ottawa are intermediate (0.128 to 0.134), as is Wolfville (0.132). That the western B. a. pooled 2 reference sample hemiophrys pooled 2 reference sample americanus Figure 16. Histograms of individual discriminant function scores for pooled reference samples of females of B. a. hemiophrys and B. a. americanus. Vertical axis shows number of individuals and the horizontal gives the score. List of localities, dates, catalogue numbers and number of specimens is given in Appendix I. americanus have smaller dorsal spots than B. a. hemiophrys heightens the contrast between the taxa in adjacent western Ontario and eastern Manitoba. Posterior Width of the Cranial Crest/Head Width: B. a. hemiophrys population means of this character range from 0.240 to 0.281. There does not appear to be any obvious geographic pattern to the ratio within the taxa, as both high and low means occur generally throughout the range, ex- cept that the two highest values (0.280, 0.281) are in central Manitoba whereas all other values are 0.269 or below. In B. a. americanus the values for means vary from 0.339 to 0.442. The low value is from Ox- drift in western Ontario. The next lowest value is 0.370 from Whitetop Creek. All other values are 0.379 (Cap-des-Rosiers, Moosonee) or above. Values at 0.398 or above are predominant in the east (six out of eight means). In extreme southern Ontario values are 0.408 at Rondeau and 0.401 at Point Pelee. The other central values are 0.390-0.395 in the Kenogami to Ottawa area and 0.382 at Lac Attila. There is no clear north-south cline. Spade/Tarsus: The mean values for B. a. hemiophrys populations varied from 0.352 to 0.402. The lowest is from The Pas, Manitoba, and the highest is from Strathmore in western Alber- ta. In B. a. americanus means varied from 0.267 at Lac Attila to 0.314 at Moncton, New 28 Brunswick, and Sept Iles, Quebec. No clear geographic trends are evident in either taxon. c. Discriminant Function Analysis of Females The discriminant weights and means for each character for females are given in Table 4. Posterior cranial crest width made the largest contribution as it did in the male analysis. Tarsus length and parotoid separation also make con- tributions but the largest wart on the right tibia, the cranial crest length (right side) and distance from nostril to parotoid on the left side seem to make disproportionate contributions to this separation compared to their contribution in the discriminant analysis of males. The Mahalanobis generalized distance is 104.6573390 for female reference samples compared to less than half that value (51.3517512; Table 1) in the male analysis. The histograms of the individual scores for the female reference sample are plotted in Figure 16. Because of the generally higher weights in the discrimination and larger values of individual measurements the scores are higher than obtained in the male discrimination. As individual popula- tion samples were small, these have not been plot- ted separately. 3. Morphological Character Comparison of Southeastern Manitoba Bufo. a. Discriminant Scores of Males Figure 3 gives the distribution of samples in eastern Manitoba. None of these collections were Table 4. Discriminant weights and mean values for each character in discriminant analysis of female reference samples of Bufo a. americanus and B. a. hemiophrys, and the weight contribution of the Mahalanobis generalized distance. Discriminant Weights Snout-vent length + 0.05570437 Nostril separation — 12.574883 Cranial crest length (right) + 8.7230959 Cranial crest length (left) + 0.99025917 Nostril to parotoid (right) + 2.4354296 Nostril to parotoid (left) — 0.99551696 Cranial crest width (anterior) — 5.6168242 Cranial crest width (posterior) + 11.437762 Eyelid length — 0.16800469 Head width — 0.66918558 Tympanum diameter — 0.59421539 Parotoid length (right) + 0.41642243 Parotoid length (left) + 1.6200924 Parotoid width (right) — 1.5974340 Parotoid width (left) — 3.1625957 Parotoid separation — 6.1137037 Spot length (right) Spot length (left) + 0.13028830 + 0.23475748 Spot width (right) +4.4119072 Spot width (left) — 1.2844067 Wart width (right) + 5.1356106 Wart width (left) + 3.3310843 No. warts per spot (right) + 0.66464460 No. warts per spot (left) — 1.7433395 Tibia length — 1.012804 Largest wart on tibia (right) + 6.8065100 Largest wart on tibia (left) + 0.53891832 Tarsus length +2.0391092 **Spade’”’ width — 10.362802 Inner metatarsal tubercle width Mean discriminant score (centroid): Mahalonobis generalized distance + 0.42757177 104.6573390 included in the reference samples used for the discriminations between B. a. hemiophrys and B. a. americanus. These collections can be divided roughly into six west-east transects for discussion: 1. The northern transect: localities 1 to 14. 2. The Beausejour transect: localities 15 to 21. 3. The Vivian transect: localities 22 to 38. 4. The Trans-Canada Highway (Highway 1): localities 39 to 43. 5. The Marchand transect: localities 44 to 48. 6. The southern transect: localities 49 to 61. Sample sizes are given in Table 5. There is great disparity in number (2-268 individuals) largely due to local weather conditions during the collecting period. Some localities represent pooled adjacent 1 29 Means Bufo a. Bufo a. americanus hemiophrys Contribution 72.499863 59.72431 + 0.711654 4.5457535 4.3228445 — 2.803055 14.881289 11.388057 + 30.471798 14.857541 11.420262 + 3.403797 15.081285 12.211794 + 6.988443 15.127056 12.233808 — 2.880277 5.8474321 5.2152014 — 3.551129 8.1440296 4.3813000 + 43.037206 8.6168976 7.2346783 — 0.232219 20.243988 16.905833 — 2.233842 5.7406559 4.3101177 — 0.850048 13.593149 10.921095 + 1.112703 13.721965 11.038900 + 4.346813 7.2253618 6.5355158 — 1.101983 7.1067219 6.5041580 — 1.905666 9.1507912 9.7041588 + 3.383126 10.289782 9.6245070 + 0.086678 10.540627 9.6906052 + 0.199549 5.2016764 4.5041809 + 3.077285 5.1830263 4.4414635 — 0.952468 3.2084694 2.0423670 + 5.988648 3.1898260 2.0398245 + 3.830752 2.6440678 4.4491520 — 1.199739 2.6779652 4.3135586 + 2.851395 26.891434 20.971939 — §.995288 3.7305031 2.1245699 + 10.930800 4.4559221 2.1906691 + 1.220786 18.149078 13.522799 + 9.433488 5.6287861 5.3338470 — 3.056395 3.0406723 2.2347364 + 0.344595 17.73267 13.075331 samples, several samples from separate visits dur- ing the same breeding season and/or several col- lections taken in different years (see Appendix I). Table 5 gives the mean discriminant score for each population sample. That western scores are low and eastern ones high is immediately ap- parent. The rapid transition from minus to plus (between — 10 and +10) scores occurs largely in the eastern half of the Manitoba Lowlands Forest Region (of Rowe 1959), with the southern transect transition in the western extreme of the Great Lakes Forest. Table 5 gives the variance values for these samples. Generally a zone of very high variances (in the 80-258 range) clearly coincides with the intermediate means in five of the 8£0°0 + 6IE 0 ¥Z0'0 + H9E'0 970'0 + PET'O T10';0+ PLe'0 VS + L°09 60+*0°S EO0+67 TCOFOE oll ME bb gs 87 €cO0'O0F VIE'O ECO OFELE 0 €70'0 + 9ET 0 v10'0 + 89£€°0 ty Eco 8'0+0°S 0'0+*0°C 0'0+0°E oe VSir ve Le c70'0 + CEO LE0'0 + 69£'°0 vr0'0 +9710 010°0 + OSE'0 © S09 SOFTS vO0+F8'T O0'OF0O'E oc =F EIT 14 9¢ C70'0 + OTE ¥7Z0'0 + H9E'0 ¢€0°0 + 801°0 v10'0 + 6SE°0 La re oS O1I+S'P v0F8'T POs toC -¢ bit 60 Sc 610°0 + €7E'0 870'0 + 09£ 0 yeo'0+ 6110 7100 + OLE'0 Perey <9 O1+0°S to -=6rl 00+F0'°€ yes “Ee Ort+ oS VC ¢7Z0'0 + IZE0 Iv0'0 + SSt°0 €70°0 + 8710 T10°0 + C9€°0 ORL ES Oo OFc-7 v0+6'1 £O0a6<¢ c Ole i + 67 tc 9700 + SPE'0 9r0'0 + SOE'0 L€0°0 + 9ST'0 €10'0 + 79E'0 Le Pop 60+9°7 SO0F9'T Oe Lot o°98 40°20 — 8e (U6 JIISUBI) UBIATA 1Z0'0 + 9TE0 8£0°0 + 99£°0 S00 + €CI'0 v10'0 + 69£°0 CEFOTS 60F1'S VO0+F8'T OO OF CYS sO et +: el 17 6100+ 9I€'0 870°0 + SLE'0 970°0 + IPI'0 LI0';0+ €LE'0 es ig al I SO 1S CO Oe 00F0' 9 Cr “ae. Sr 8V 0c €£0°0 + 80€°0 9700 + €8€°0 0v0°0 + I71'0 O10'0 + 890 LYyroes vos TS vO+6T v0+67 Soc le0°ot + ba 61 ¢7Z0'0 + 9PE'0 9€0°0 + 067'0 €£0°O0+ ILT'0 T10°'0 + p9€'0 ot +265 (ag ma Nag $0.at*t L Oa 91 0'76 0 :10- oT 81 €70'0 + EPE'0 620'0 + 1870 c£0'0 + H9T'O c10'0 + S9E'0 Scrs OS OT. ey 90 Fc | L Orr Gal Loe oo Ol = OL LI 610°0 + 99€°0 870'0 + 1670 $£0°0 + 09T 0 O1TO'O + £9E.0 6c +P OF + Lor SOT v0+60 CUC a St = lin 91 070°0 + 98€°0 810°0 + $87:0 €c0'0 + LOT'0 800°0 + 9S¢°0 Lab + o°CS O1+t'P 9:0 a2) 00 OF VEC Us Gl el Si posuvd) mnofasnveg 870'0 + 9870 100°0 + I18€°0 870'0 + ZOI'0 T10'°0 + LSE°0 8°0 + £°89 0'0+0'S O0FO-¢ 0 0s Ost S60'es Si (6 vi 8£0°0 + E7E'0 800°0 + 8£€°0 IS0°0 + OIT'0 900°0 + 79€°0 61 +0°P9 0';0+0°S OO =OT 00+F0'E 96¢c 8 S0+ € el 170'0 + OIE 0 870'0 + O9£ 0 6£0°0 + 8IT 0 110°0 + 09€°0 LoS 09 OT+CY yO+8' T 0O0F0'E' ViGY 8 Tt + SI cl 8100+ ££ 0 1v0'0 + L9E°0 $L0°0 + Prl 0 T10'0 + V9E'0 ESTE CS US ey 1h Oe os ea br OrG.c CUD ook Olt cr IT 810°0 + 97£'0 c£0'0 + 9VE'0 €£0°0 +0010 C100 + PSE°0 Pero 19 60+LP vOr+8 I EOtG ¢ i me a8 a Ie Ol $S0'0 + €7£°0 €£0°0 + v67 0 Tv0'0 + O9T'O e100 + 89€°0 EC Like 80797 SO eT £O+1 1 "0c FOr 81 6 LOO CSE 870'0 + 0670 COD FE! 0 LlOOF COE 0 (OR doce I 4 9 EL or 90+0'1 SO 971 8€¢ “P60 - VC 8 870'0 + IPe'0 970'0 + PL7'0 L700 + ITO ClO OF 99¢°0 CoE F Saks CLO e $'0+9°0 POF oT CBE 20k (44 ye 0€0°0 + OFE'0 vt0'0 + 6L7'0 0£0'0 + SET 0 vI0'0 + OLE'0 OV +7'6S S0+0'P 8'0FPr'l SOs S23 tly of £0— O€ 9 7700 + EVE'0 9£€0°0 + 067°0 $t0'0+ EFI 0 010°0 + 890 dae 607 SOtL | OU Ls Cio 5 30 I€ g 0v0'0 + S9E°0 cv0'0 + L670 8£0°0 + OLT'0 010'°0 + S9E°0 Lito Oe Lee Tay J 9 Mc | S07 6.) 6.09 ay Cl — SI v 910°0 + 9FE'0 c£0'0 + €87°0 LeQO* £97.0 T10'0 + p90 87 +7 6S t Or ES LO rel v OF CT O5EC Gel = el 3 £60°0 + 88€°0 ep°0 = .120:0 6720°0 + S9T'0 980°0 + 8r£'0 VCTHB'PS b leroy 90-410 v0+0'T OY 0 SI ¢ vZ0'0 + 19€°0 L700 + €87°0 v£0'0+ CLIO 600°0 + 6S£°0 O'V+S'6S 01+ P 9°00" tOF0T Sus 6.0C = te I JIISUBI} UIIYJION Sx SEN Sx SFX SFX S#xXx Sra X SFX 7s x u Jaquinu ouel onel one one y13sud] 3109S 31098 3109S 3109S 9ZIS AN][@90 7 yisuay/yIpIM SUIPIM YIsUud] Ysud] USA IOUS A, [e1qio (99) 183.19 uonouny ajduies snsie1/apeds pesy/oo AS/10dS AS /PIQLL -jnous -1SOd jetuelg JUBUTUILIOSIG, 30 ‘(SUNOS I9YyIO [][e) SF ‘UOTIBIAIP piepuels 9U0 F JO “(3109S UONSUN] JUVUTUILIDSIP) 7S ‘9OULTIVA SY] YIM UDAIS SI s[dures YoRd IO} ‘x ‘UBS OY “JAS JaJODUDYI jVI180;0 Yao JOPUN UONIES SpOYJafA] PUB [PIIOIVIY 94) UI Paqliosap pue ¢ pUe fp SOINSI Ul UMOYS de pasn sidjoVIeVYD ‘“UOTIDa][OO JO salep pue sJaquINU onsoyejyeo YM | xipusddy ur pojsi] oie pue ¢ dinSIy ul poddew oie sarjyeoo] ayy, ‘eqowUR UJoIseayINos WoL Ofng Jo (sayewW C9PZ) so]dwies [9 BUI[eIO} SJOISUBI] XIS JO UOSTIEAWIOD OJ SONBI Pd}daJTaS INO] pUL JUSWAINSeaU DUO ‘sdIODS ddIY] ‘UOTIOUNJ JUPUTULIOSIG “Ss aIquy 1Z0°0 + 90£°0 8Z0°0 + S9E°0 LE0'0 + 8IT 0 v10'0 + S9E°0 Ltr Pals 607S'°P OF 6.7 0'0+0'€ VLoe Seed 89 19 070°0 + L670 €80°0 + LLE°0 ye0'0+SII'0 110°0 + I9€°0 L't'* €-09 O1+S'P v0+F6'1 0'0+F0'E Soo i-sicr 69 09 0€0°0 + ITE0 870°0 + L9E0 tv0'0 + THI 0 710'0 + I9E°0 6E+9'19 8'0+S'P £O+61 Con ¢ O6E Ccls St 6$ 1v0'0 + SOE°0 0£0°0 + ZLE'0 8£0°0 + S710 810°0 + 0LE'0 0’ +709 O1+8'P €'0+F6'1 COUTEE cot: Ger ta Sc 8S 870'0 + brE'0 8£0'0+ PIE 0 670°0 + 7910 7100 + £9€°0 [vr CES SI FCE LE Oa $1 SO Fe ft 879 S$ sO- £9 LS 87Z0°0 + 10£°0 6£0°0 + SPE 0 $£0°0 + S710 110°0 + 6LE°0 Creer Bey ie ee c07O0%¢ S'OF9'C Sse Scio 0¢ 9¢ 770'0 + LZE0 £r0°0 + 9EE°0 8£0°0 + 99T°0 €10°0 + CLE'0 O'S.* Capo 8'OFS'P v 0F 6 SOTSC t-ccl Ge0— vl $s 9700 + SPE'0 se0°'0+ €1E°0 8£0°0 + 7rI'0 T10°0 + 69£°0 67 +£°09 CAL eP Ff. 0s S71 8°0+0°7 L1OL Evo — 87 vs 970°0 + Lre'0 9£0°0 + L670 T10'0 + 9LT 0 €10'0 + $9€°0 6c 006 O1+8'P 50 9°] S.0 ap I COs: Gcl= 6 €s £€70'0 + 6hE°0 8700 + 7670 $£0'0 + €LI°0 7100 + H9E 0 Cts ass Ot HPF SOF S01 SO+FS'T Gye” Sol Te SV ge 670'0 + PSE0 970°0 + 7670 970°0 + 691°0 I10°0 + SSE°0 LO GSE L029 ¥. LO Col v'0+8°0 aoc, Cathe SS és €£0'0 + ILE'0 6720°0 + €L7°0 0£0°0 + L9T'0 7100 + 9PE'0 ¥CTS CP O1+9'P L001 v'0+8°0 GLI wrt Lt OSs €70'0 + 7SE'0 0£0°0 + 9L7'0 L70'0 + 6S 10 I10°0 + S9€°0 O’e + L288 8 O+FL'P S0+L0 G0 +t I ett. Stl vv 60 }9SUBI} UJIYINOS ¥Z0'0 + 60£ 0 8700 + £9£°0 9r0°0+ LITO 600°0 + 6S€°0 0'r+0'19 6087 C0707 0O0F0'E' Vso. LO ler or 8V 670'0 + OI€'0 0£0°0 + P8E°0 Lv0'0 + SIT°0 7100 + 19€°0 V€+8°6S SOF6? t 04 07 00 F0°S se. Glo we LV 1€0°0 + 8ZE'0 €70°0 + L9E"0 c£0'0 + OFT 0 €10'0 + S9E°0 S Ct E99 8067 00+0°7% SoTL Ose. EO Sc 4 770'0 + PLEO £700 + 667'0 910°0 + 981°0 L00°0 + 79£°0 GEA Eo vOFTS SOF Lol Voc I OC Sr te “ SY 970'0 + PrE'O ¥7Z0'0 + 9L7°0 $£0°0 + OLT'0 I10°0 + @SE°0 6CF9OSP OF 0 Y FO0tCT v'0+8°0 bbl Sev le 98 vv posuvl) puvydieyy €70' 0+ FIE'O £90°0 + 6LE°0 670°0 + IZI'0 I10'0 + SSE°0 Ler 65 90+F6'P €0+6'1 00+0'E 8:VE< 9.0 Lar orl ey 110°0 + 8ZE°0 $£0°0 + €8£°0 €v0'0+SIT'0 110°0 + 79€°0 Ly+9'y9 SOFP'S e067] 0'O0+0'E O0E Felt L (44 610°0 + £0£°0 670'0 + €LE°0 Lv0'0 + I7I0 €10°0 + 99€°0 Cr LCS LOTS 0';0+0°C OO+F0E L bts SoS jb Iv €70'0 + FIE'O 1€0°0 + 09€°0 ye0'0+6IT'0 S10'0 + SSE°0 Vt Sco 6067 © OF G6 OOFOE Eve. SOL Se OV €70'OF ITEO $70'0 + 6LE°0 6£0°0 + LITO 110°0 + 9SE°0 6°€ + €°09 L0+6? PO 07 cOFOE ONG. Carl r vLi 6€ 910°0 + 80£°0 0r0'0 + 69€°0 €c0'0 + £91 °0 7000 + P9€°0 L'1+8°6S 90F9'°S S081 OO -0'C Ptcc £601 v 8E 610°0 + 97E°0 0v0'0 + S9E°0 1v0'0 + STT°O L00°0 + OSE°0 OV F919 SOFES t OF 64 0'O0+F0'€ soo Otc 6 LE 810°0 + 90€°0 L700 + L9E°0 870'0 + £710 010°0 + 7SE'0 Lc Se SOFI'S €0+6'1 0'0+0'E V3 6.0L7 VC of ££0°0 + 07E'0 9£0°0 + 8h£°0 $v0'0 + OFT'0 vI0'0 + LSE'0 Cv +0'6S SOL? vO0+8'T 910 5°C $68 psot «897 St 1v0'0 + O£E°0 yr0'0 + I€t°0 8£0°0 + Srl0 810°0 + 09E°0 8°€+O'8S 60+8°P yO0+8'T 80FS'T 8°46 810+ Le ve 770'0 + 99€°0 8£0°0 + 167°0 c£0'°0 + P9T 0 vI0'0+ EEO L€+O'PS S0+6'V SOFT SOFT Ese Sele oP cf €70'0 + ESE°0 £€£0°0 + 887'0 c70'0 + OLT'O €c0'0 + CSE'0 St ELS bol Pp 9O0F LI Piet | COG. are ts 0c cE €70'0 + 6SE°0 0£0°0 + 997°0 €£0'0+ 79T'0 6700 + TSE'0 Ltr evs SOFL F eS oO Fah I Lies Gok evi It 070°0 + O8€°0 vr0'0 + L67°0 870'0 + SST°0 O10'0 + SSE°0 £ LiF 9s OLECP 50+ £0 v'0+8°0 0:03" Col> 9 O€ 7700 + TSE0 67200 + 0970 L70'0 + 8L1°0 O10°'0 + PSE'0 VTS PS LOF9'P DOrReO-T v'0+6°0 Voces (aS be LE 67 yosuvl) ABVMYSIY BEpeURD-suBIT SFX $s > Xx a es SFX S30 Sx S#xX SX 2s ¢ u Joquinu one one one one yisud] 3109S 31098 a109s 3109S dZIS Alyeoo 7 yisus,/yIpIM SUIPIM ylsud] yisud] JUSA I9]Ud A, [e1gio (99) 1s919 uonounj ajduies snsie1/apeds pesy/oo AS/10dS AS /BIQLL -1noUus -180d jeturlg JUBUILWULIOSIG ‘(| xipuaddy as) suautdads plZ UO paseq sorsnejis ONLI pue ajelIeAIUN ‘sudUTIOads ggZ UO paseq UONOUNJ JUPUIUITIOSIC 31 +40 S66) % 84 So. 2 O oO +20 A A AO A Oo A x - 4 ri “a Loa a, A A 4 A 4 ws +10 a = o Be O A A WY ke A z oO 0.1 Intermediate pop. 62 — 0.355 <0.01 B. hemiophrys 111 — 0.304 <0.01 C) Correlation with Pulse Rate B. a. americanus 50 Intermediate pop. 62 B. a. hemiophrys 111 D) Correlation between Call Length and Dominant Frequency B. a. americanus 50 Intermediate pop. 62 B. a. hemiophrys 111 Dominant Call Length Frequency r P r P — 0.635 <0.001 + 0.416 <0.01 — 0.359 <0.05>0.01 + 0.272 <0.05>0.01 —0.459 <0.001 — 0.036 >>0.1 —0.170 >>001 — 0.196 >S0. 1 — 0.267 <0.05 — 0.183 >0.1 — 0.018 S> 0.1! — 0.246 <0.05 r P r PF — 0.653 < <0.001 + 0.409 <0.01 >0.001 — 0.588 < <0.001 — 0.034 > >0,1 — 0.499 < <0.001 — 0.027 > >0.1 r P — 0.235 <0.0 + 0.130 >> 1 + 0.031 >>0i1 Table 7. Discriminant weights and the weight contribution to the Mahalanobis generalized distance from the discriminant analysis of one morphological (snout-vent length), one physiological (body temperature) and three call (pulse rate, dominant frequency and duration) parameters for Bufo a. hemiophrys and B. a. americanus. Discriminant Character Weight Snout-vent length + 0.95150 Body temperature + 3.42561 Pulse rate — 1.23113 Dominant frequency + 0.00477 Duration + 0.92707 Mean discriminant score: Mahalanobis generalized distance: 39.166 Table 6 gives the results of correlation tests be- tween each possible pair of variables. The number of pulses per second increased directly with an increase in temperature in both B. a. americanus and B. a. hemiophrys samples as shown by the correlation coefficients (0.959 and 0.929 respectively). The relationship of tempera- Mean for Mean for Weight B. a. americanus B. a. hemiophrys Contribution 60.216 55.181 + 4.857 14.989 19.440 — 15.248 27.644 74.622 + 57.836 1871.340 1847.407 +0.114 1B be 3.448 + 7.160 93.885 54.719 ture and call length is inverse and the correlation is weaker (— 0.635 and — 0.459 respectively) than with pulse rate. There is considerable variation from call to call within a single individual at any given temperature and the value used here is an average of several calls from each individual. The relationship of dominant frequency to temperature 47 HEMIOPHRYS REFERENCE SAMPLES AMERICANUS INTERMEDIATES HWY | TRANSECT NIVERVILLE 31.25 - 29.75 mi. W. Figure 31. Histograms of call discriminant scores for individual Bufo in reference samples and in eight intermediate populations. is positive and the correlation, although signifi- cant, is not high in B. a. americanus (+ 0.416, P<0Q.01) but negative and not significant in B. a. hemiophrys (-—0.036, P>>0.1). In- termediate populations show weaker correlation between pulse rate (+ 0.583, P<0.001) and call length (— 0.359, P<0.05 >0.01) with temperature, and are intermediate in the relationship of domi- nant frequency and temperature (+0.272, P<0.05>0.01). Weak positive correlation was present in B. a. hemiophrys between snout-vent and pulse rate 48 (+ 0.304, P<0.01) but the correlation in B. a. americanus was not significant. In the inter- mediate populations the correlation was negative and weak (— 0.335, P<0.01). Call length showed a weak correlation with snout-vent length in in- termediates (+ 0.267, P<0.05) but was not signifi- cant in B. a. americanus or B. a. hemiophrys. Dominant frequency showed weak negative cor- relation and significance (— 0.246, P<0.05) in B. a. hemiophrys but was not significant in B. a. americanus or intermediates. Note that dominant frequency was not correlated with temperature, wi x e) 2) n = = < z = x 2) 2) @) =i = < Oo O HEMIOPHRYS A INTERMEDIATES O AMERICANUS 10 MORHOLOGICAL DISCRIMINANT SCORE Figure 32. Relationship of call discriminant score to morphological discriminant score in three samples of Bufo (Appendix II gives localities and number of specimens). but was weakly correlated with size, in B. a. hemiophrys. Pulse rate correlation with call length was negative and weak but significant. Dominant fre- quency showed a weak positive correlation with pulse rate in B. a. americanus and an almost negligible negative correlation in intermediates, but was not significant in B. a. hemiophrys. Call length in B. a. americanus showed a weak negative correlation with dominant frequency (—0.235, P<0.01) in B. a. americanus but no significant correlation in intermediates or B. a. hemiophrys. In B. a. hemiophrys in this study snout-vent length correlations with pulse rate (— 0.304) and dominant frequency (— 0.246) were significant (P <0.01, P<0.05 respectively). 49 Discriminant Analysis of Call Variables and Com- parison with Morphological Discriminant Scores: A discriminant function using the three call variables (pulse rate, dominant frequency, and duration) as well as temperature and snout-vent length compared equal B. a. hemiophrys and B. a. americanus samples of 45 specimens each (see Appendix II). This analysis produced completed separation of the two taxa. Application of the resultant weights (Table 7) to the individuals from geographically inter- mediate samples gave intermediate values (Figure 31). Pulse rate made the largest contribution to the separation between means (centroids). Figure 32 plots the call discriminant vs. the mor- phological discriminant values for B. a. ameri- canus, B. a. hemiophrys and intermediates. Table 8 gives results of correlation for each of the three units and the total sample. There is a positive cor- Table 8. Correlation (7) and significance (P) values for comparison of morphological discriminant function and call discriminant function. No. sp. r P Bufo americanus 50 + 0.188 > >0i Intermediate populations 62 + 0.552 < <0.001 Bufo hemiophrys 111 — 0.178 <0.1>0.05 Pooled sample Zea + 0.820 < <0.001 Table 9. Results of cross-mating and artificial rearing: number of individuals in each sample (50 eggs) which reached transformation. Number reaching transformation Cross Laboratory 17 miles [27.4 km] west 31 miles [49.9 km] west AA-1 (May 7) 45 (July 7-19) 11 (July 12-14) 11 (July 12) AA-2 (May 16) 42 (July 8-Sept. 13) 4 (July 12) 22 (July 13-14) AA-4 (May 27) 17 (July 14-24) . = AA-S (May 27) 44 (July 14-Aug. 25) 26 (July 14-16) * AI-4 (May 27) AH-1 (May 16) AH-4 (May 27) IA-5 (May 28) II-1 (May 25) II-4 (May 27) II-S (May 27) II-6 (May 28) II-7 (May 29) IH-1 (May 16) IH-4 (May 27) IH-5 (May 27) IH-6 (May 28) IH-7 (May 28) HA-1 (May 16) HA-2 (May 16) HA-3 (May 16) HI-2 (May 16) HI-3 (May 16) HH-4 (May 27) HH-S5 (May 27) 40 (July 14-Dec. 13) 47 (July 4-Aug. 11) 18 (July 26-Oct. 2) 23 (July 15-Aug. 9) 39 (July 3-Oct. 5) 15 (July 11-Sept. 1) 3 (July 16-28) 37 (July 13-Aug. 31) 32 (July 18-Sept. 29) 11 (July 7-14) 2 (July 11) 18 (July 14-Aug. 25) 21 (July 14-Sept. 3) 26 (July 18-Nov. 12) 3 (July 7-Aug. 13) 12 (July 9-19) 8 (June 30-July 12) 11 (July 10-30) 7 (July 4-12) 6 (July 17-Aug. 25) 12 (July 19-Sept. 16) 7 (July 15-16) 24 (July 13-16) 17 (July 15-19) 17 (July 14) 10 (July 14) 7 (July 13-16) 4 (July 15-16) 15 (July 14-16) 10 (July 14-16) 16 (July 12) 18 (July 13-29) 9 (July 14-16) 15 (July 14-17) 16 (July 14-16) 14 (July 13-21) 12 (July 12) 13 (July 13-14) 7 (July 14-16) 5 (July 12) 6 (July 14-15) 15 (July 14-21) 5 (July 14) 20 (July 13-18) 15 (July 13-17) 8 (July 14-Aug. 8) 20 (July 13-17) * 10 (July 14-19) 22 (July 14-29) 28 (July 15-Aug. 7) 23 (July 14-25) 4 (July 14-16) 4 (July 13-14) 25 (July 13-15) 4 (July 14) 13 (July 14-26) 26 (July 13-15) 23 (July 13-16) 24 (July 13-15) 25 (July 13-21) 25 (July 14-31) 8 (July 15-29) A B. a. americanus parent from 3 miles [4.8 km] west on Highway 1 of junction of Highways 1 and 11. H_ B. a. hemiophrys parent from 31-29 miles [49.9-46.7 km] west on Highway 1 of junction of Highways 1 and 11. I intermediate parent from 17 miles [27.4 km] west on Highway 1 of junction of Highways 1 and 11. In each combination the first letter designates the female parent, the second the male; the number following the dash is the sequential number assigned a replicate within that cross combination. Missing numbers in sequence are crosses which failed to produce fertile eggs. Dates in parenthesis are: the date cross was made (after cross number) and dates of first and last in- dividual to transform (after number reaching transformation). * Indicates a prehatching failure at a particular locality due to accident or obvious contamination (e.g. fungal growth). In Figure 33 any replicate with a failure in laboratory or field is omitted in computing the mean. 50 relation (+0.820, P< <0.001) within the whole sample, and a correlation (+ 0.552, P< <0.001) for morphology and call within the intermediate populations. However, within B. a. hemiophrys or B. a. americanus there is no significant correla- tion between call and morphology, indicating they are varying independently within each taxon. These correlations (of total sample and inter- mediates) are evident despite the fact that a por- tion of the reference sample for call data includes populations where some introgression is evident in morphological characters, as demonstrated earlier. 5. Crosses Reared in Laboratory and Field During the 1969 field season, males and females were selected from three sites along the Trans- Canada Highway: west of its junction with Highway 11: 31-29 miles [49.9-46.7 km] (B. a. hemiophrys), 17 miles [27.4 km] west (inter- mediate) and 3 miles [4.8 km] west (B. a. ameri- canus). Crosses were attempted of every possible combination of males and females among these three populations (see Materials and Methods). In all, 61 pairings were tried but only 28 of these pro- duced eggs, and in three of these the eggs did not Mean Number Transformed Number of Replicates N Parents (female; male) A- americanus |- intermediate H - hemiophrys develop, leaving only 25 successful crosses. However, they included at least one cross of every possible combination. The number of individuals metamorphosing in each cross, the date the cross was made, and the dates of first and last transformation are given in Table 9. The mean number of individuals trans- forming from all complete replicates (with some transforming individuals in laboratory and both ponds) are compared in Figure 33. In the laboratory, crosses using B. a. americanus females had high transforming suc- cess, and those using B. a. hemiophrys females had low success, with intermediate females yielding intermediate results. The water used in the laboratory was pumped in from Shoal Lake (per- sonal communication, K.W. Stewart) in eastern Manitoba, an area well within B. a. americanus territory. No similar marked contrasts existed within or between the field-raised replicates. A Chi-square test of pure B. a. americanus and pure B. a. hemiophrys between intermediate (17 miles [27.4 km] w.) and B. a. hemiophrys (31 miles [49.0 km] w.) ponds gives a value of x7=1.30, which is not significant (0.5 >p>0.1). The experiment lacks sufficient successful replicates, is unbalanced Figure 33. Histogram of mean number of tadpoles reaching transformation (from samples of 50 eggs) for each combination of B. a. americanus, intermediate, and B. a. hemiophrys parents in laboratory (open bars), in pond 17 miles [27.4 km] W on Highway 1 of junction of Highways 1 and 11 (horizontally hatched bars), and in pond 31 miles [49.9 km] W on Highway 1 of junction of Highways 1 and 11 (vertically hatched bars). The first letter given of each cross refers to the female parent, the second is the male. = or) G ‘ NX \ \ Bufo boreas NMNS(8)<20 (©)>20; UAA Bufo a. hemiophrys NMNS(H)<20 (h)>20; UAA Mixed population NMNS(]); UA 4\ Vegetation zones Aspen Parkland Boreal Forest Figure 34. Map of west central Alberta showing the distributional relationship of Bufo boreas (B,b) and B. a. hemiophrys (H,h) in a narrow zone of sympatry. A capital letter indicates more than 20 males collected, a lower case letter, fewer than 20 males. All collections from breeding choruses. UA refers to collections in the University of Alberta, and NMNS to those in the National Museum of Natural Sciences. because of unequal replicates between crosses, and shows wide variation in transforming success bet- ween some replicates (a difference of as much as 28 individuals transformed). In addition, some laboratory crosses took an abnormally long time to transform (e.g. to December 13 in AI-4!), possibly because of suboptimal temperatures and/or food. 6. Comparisons of B. a. hemiophrys and B. a. americanus with Other Bufo in Northern North America a) B. a. hemiophrys-B. boreas: In 1965, the area north of Edmonton was examined during peak toad breeding and four breeding sites were found where B. boreas and B. a. hemiophrys were pre- sent in mixed choruses. Sites visited to the east of these contained only B. a. hemiophrys and those to the west only B. boreas (see Appendix I, Figure 34 and Table 10 for localities). Examination of the collection in the University of Alberta Zoology Department documented a slightly more extensive overlap zone (Figure 34). 52 In two of the mixed chorus sites examined, B. boreas was the more common species (but one of these contained only three calling males), in the other two B. a. hemiophrys predominated. Mismated pairs were found, usually a female of the less abundant species mated with a male of the predominant species. Only one individual suspected of being a hybrid was found, that ina pond (2.4 miles [3.9 km] north of Westlock) where B. a. hemiophrys was the more common species. To compare the two species and evaluate the suspected hybrid, a discriminant comparison was made between 48 B. a. hemiophrys (NMC 8521) and 60 B. boreas (NMC 8530). The B. a. hemiophrys were from the same pond where the hybrid was taken, but because of the smaller sample of B. boreas available from this locality, a larger series from a nearby locality (2 miles [3.2 km] west on Highway 2 of Athabaska) was used for the reference sample of that taxon. Measurements were those taken for the B. a. hemiophrys-americanus comparison, except that all characters involving the cranial crest had to be Table 10. Summary of localities where Bufo boreas and B. a. hemiophrys were found using the same breeding ponds in west-central Alberta during the 1965 field survey of the area. 24 May 1965 3 miles [4.8 km] W on Highway 18 of Freedom (turnoff) 26 May 1965 2.4 miles [3.9 km] N on Highway 44 of Westlock 27 May 1965 0.6 miles [1.0 km] SW on Highway 2 of Perryvale (turnoff) 28 May 1965 2 miles [3.2 km] W on Highway 2 of Athabasca B. boreas males females 2 0 7 0 8 1 60 5 B. a. hemiophrys males females l 0 48 15 36 0 4 1 Hybrid male Table 11. Discriminant weights and weight contributions to Mahalanobis generalized distance for each character for B. a. hemiophrys and B. boreas, mean discriminant scores for both species and values for a hybrid individual. Discriminant Weights : Snout-vent length — 4.842178 Nostril separation + 29.637020 Nostril to parotoid (right) + 2.362891 Nostril to parotoid (left) —0.9124445 Eyelid length + 1.16125 Head width + 0.1174592 Tympanum diameter + 11.6957 Parotoid length (right) + 4.602726 Parotoid length (left) — 4.331665 Parotoid width (right) + 1.079502 Parotoid width (left) + 2.935235 Parotoid separation + 2.59703 Spot length (right) + 1.009183 Spot length (left) — 1.57778 Spot width (right) + 0.3525454 Spot width (left) + 1.497334 Wart width (right) — 2.970933 Wart width (left) — 2.052994 No. warts per spot (right) + 0.05815297 No. warts per spot (left) + 1.092478 Tibia length + 5.659762 Largest wart on tibia (right) — 1.178755 Largest wart on tibia (left) — 1.918204 Tarsus length — 4.341073 **Spade’’ width + 21.52207 Inner metatarsal tubercle width —16.23179 Mean discriminant score (centroid): Mahalanobis generalized distance: 126.42582 Mean for B. a. hemiophrys NMC 8521 60.46243 4.589577 1P-87079 11.84996 7.347894 17403535 4.395826 10.9812 11.164530 6.929142 6.924962 9.358281 10.19787 10.19995 4.904161 4.783328 2.204161 2.22916 5.1875 4.791666 22.22492 Dial 2912, 2.164577 14.23952 5.177079 2.208327 + 106.9462 53 Mean for B. boreas NMC 8530 62.88822 3.9566610 10.88994 10.93160 7.429959 7.29992 3.734993 11.06492 11.10326 7.296618 7.326615 8.798274 9.464948 9.453274 3.888327 4.029994 3.179994 3.373328 3.216666 3.259592 ASST 11.67494 11.913280 15.18993 3.496661 2.686661 19.47962 NMC 8523 74.4 13.8 13.8 19.2 13.1 12.6 ee 14.0 12.5 25.3 oa OH 16.9 a3 Z.3 + 63.0288 Weight Contribution + 11.74610697 + 18.757774415 + 2.31641637 — 0 B35 795Z534 — 0.095297981 — 0.031076181 + 7.728904518 — 0.385340221 — 0.265401115 — 0.396691077 — 1.178945943 + 1.454354979 + 0.739652423 — 1.178090459 + 0.358127604 + 1.127992612 + 2.899134462 + 2.348970039 + 0.114609851 + 1.702445612 + 1.660291183 + 11.08268752 + 18.70000109 + 4.125799190 + 36.16607383 + 7.764217038 boreas hemiophrys Figure 35. Histograms of discriminant function scores for reference samples of male Bufo boreas and B. a. hemiophrys and one natural hybrid between these taxa. Table 12. Discriminant weights, weight contributions to Mahalanobis generalized distance, and means for each character for discriminant function analysis of Bufo a. hemiophrys and B. cognatus. Discriminant Mean Mean B. a. Weight Character Weights B. cognatus hemiophrys Contribution Snout-vent length + 1.9618 72.3333 63.0741 + 18.165 Nostril separation — 10.0243 4.8292 4.6034 — 2.263 Cranial crest length (right) + 10.1456 13.9979 11.9741 + 20.532 Cranial crest length (left) — 0.9120 14.0417 12.0690 — 1.799 Nostril to parotoid (right) — 3.0633 14.8313 12.9345 — 5.810 Nostril to parotoid (left) — 1.2635 14.8958 12.9586 — 2.448 Cranial crest width (anterior) — 21.5565 6.2792 5.5931 — 14.789 Cranial crest width (posterior) + 17.8171 9.1292 4.6121 + 80.481 Eyelid length — 2.2461 8.5875 7.8138 — 1.738 Head width + 2.5182 21.1250 17.8052 + 8.360 Tympanum diameter + 1.1682 5.0458 4.7552 + 0.340 Parotoid length (right) — 4.1194 10.7563 11.3224 + 2.232 Parotoid length (left) + 3.2425 10.6896 11.4759 + 2.550 Parotoid width (right) — 5.9609 6.7646 7.1259 +2.154 Parotoid width (left) + 2.5110 6.6667 6.9828 — 0.794 Parotoid separation + 0.2284 13.9333 10.1948 + 0.854 Spot length (right) — 0.2942 16.4396 9.8328 — 1.944 Spot length (left) — 0.0370 14.9000 9.3328 — 0.206 Spot width (right) + 4.0041 7.6417 4.4724 + 12.690 Spot width (left) — 0.9240 7.3875 4.4052 — 2.756 Wart width (right) + 10.2861 2.0854 2.1328 — 0.487 Wart width (left) — 2.9931 2.1104 2.1138 +0.010 No. of warts per spot (right) + 0.4006 28.2083 3.500 + 9.898 No. of warts per spot (left) + 0.4401 25.7500 3.6207 + 9.740 Tibia length — 2.7946 25.6970 23.3534 — 6.552 Largest wart on tibia (right) + 0.5363 2.5042 2.2966 +0.111 Largest wart on tibia (left) + 1.5136 2.6312 2.4224 + 0.316 Tarsus length + 2.4015 15.9792 14.4948 + 3.565 ‘*Spade’’ width + 3.7639 6.3917 5.4586 + 3.512 Inner metatarsal tubercle width + 24.3866 2.9250 2.1724 + 18.353 Mean discriminant scores (centroids): 249.605 97.228 Mahalanobis generalized distance: 152.377 hemiophrys cognatus Figure 36. Histograms of discriminant scores for reference samples of male Bufo a. hemiophrys and B. cognatus. omitted since B. boreas usually lacks crests, or has crests so low that they are barely discernible and rarely can be measured with confidence. Several other characters which distinguish B. boreas and B. a. hemiophrys were not included in this com- parison. Besides the lack of cranial crests, B. boreas also lacks a dark throat in the male, and possesses a prominent tarsal fold. The discriminant weights, weight contributions to the Mahalanobis generalized distance, means for B. a. hemiophrys and B. boreas samples, and the measurements for the suspected hybrid are given in Table 11. The weight contributions indicate that ‘‘spade’’ width, nostril separation, size of tibia wart, body size, inner metatarsal tubercle width, tympanum diameter and tarsus length are making the largest contribution to separating the taxa in the analysis. B. boreas has a smaller spade, a larger inner metatarsal tubercle, a much larger tibial wart, somewhat larger snout-vent, smaller tympanum diameter and a longer tarsus. The mean of the discriminant scores for B. a. hemiophrys was + 106.9462 (+ 78.16148 to + 132.4602) and for B. boreas + 19.47962 (+ 44.11512 to — 14.2396). The suspected hybrid had a score of + 63.0288. Histograms of the discriminant scores are given in Figure 35. In addition to its intermediate score the hybrid had distinct, but open, cranial crests, with no in- dication of a B. a. hemiophrys-like central filling or posterior bridging. The anterior width of the cranial crest was 4.9 mm and the posterior 6.1 mm and on the B. a. americanus-hemiophrys boss 5 scoring system it would be assigned a score of 3, clearly distinguishing it from the 0 or 1 score nor- mal for a B. a. hemiophrys. The crest measure- ments, posterior wider than anterior, are the reverse of those typical for B. a. hemiophrys. Post- orbital crests were absent and therefore scored 0. The venter scored 5. The hybrid lacked the tarsal fold typical of B. boreas. Although the spot length does not differentiate the two species in this analysis, B. boreas often has spots that run together, in contrast to the distinct blotches typical of many B. a. hemiophrys. b) B. a. hemiophrys-B. cognatus: Only one mixed breeding chorus of B. a. hemiophrys and B. cognatus was found in the course of herpetofaunal surveys of the Canadian Prairie Provinces, but this was the only time when any breeding aggregation of B. cognatus was found during these surveys. B. cognatus breeds typically after heavy rains. On 28 May 1963 a large prairie slough 8.3 miles [13.4 km] east on Highway 1 of the Piapot turnoff yielded 48 male and 4 female B. cognatus and 22 male and 2 female B. a. hemiophrys. No mismated pairs were noted. B. cognatus is morphologically distinctive from B. a. hemiophrys. Males in the Piapot series ranged from 66.1 to 76.4 mm snout-vent length. Particularly obvious were the characteristically heavy cranial crests which unite on the nose and diverge strongly posteriorly. The postorbital crests are strongly developed, and the large, prominent dorsal blotches each contain 14 to 54 small warts. The outer metatarsal tubercle is large and broad. The venter is slightly spotted to unspotted (score values of 3(2), 2(12), and 0(34) among the males). A discriminant analysis was made between the 48 male B. cognatus from Piapot (NMC 7110) and 22 male B. a. hemiophrys from Piapot (NMC 7111) pooled with an additional 36 males from nearby Tompkins (NMC 6014). The discriminant weights, weight contributions, and mean values for each character are given in Table 12 and the discriminant scores presented as a histogram in ' Figure 36. This analysis confirms the distinc- tiveness of the two taxa and the lack of inter- mediates where they are sympatric. The largest contributions to the separation be- tween these taxa are made by the posterior cranial crest, cranial crest length, inner metatarsal tu- bercle, snout-vent length, spot width and number of warts per spot. B. cognatus is a larger toad than B. a. hemiophrys, with much greater posterior divergence of the cranial crest, larger spots with more warts per spot, and a somewhat larger spade. c) B. a. hemiophrys-B. w. woodhousei and B. a. americanus-B. w. fowleri: Two small series, one each of B. w. fowleri (19 males taken 19 June 1973 at Long Point, Norfolk County, Ontario) and B. w. woodhousei (10 males 22 June 1965 and 7 males 23 June 1965 from 3.5 miles [5.6 km] southeast of Stockton, Rooks County, Kansas) were available for comparison (see Appendix I). The weights produced by the discriminant separation of B. a. americanus and B. a. hemio- Dhrys were applied to B. w. woodhousei and B. w. fowleri specimens to determine where they would rank in- comparison with these taxa. The results are given as histograms in Figure 37. In- terestingly, they both score as intermediate be- tween B. a. hemiophrys and B. a. americanus. Both subspecies of B. woodhousei produced near- ly identical scores on the basis of the characters used. The variance in these samples, however, is low in contrast to the generally high variances of populations in the B. a. americanus and B. a. hemiophrys contact zone. Other characters, which are not included in this comparison, more-or-less separate the subspecies of B. woodhousei and more effectively separate each from B. a. ameri- canus and B. a. hemiophrys. woodhousei 3.5 mi. SSE. Stockton, Rooks Co. Kansas 8585 - 10 fowleri Long Point, Norfolk Co. Ontario 15972- 19 Figure 37. Histograms of discriminant scores for two samples of male Bufo woodhousei. Weights produced by the Bufo a. americanus-hemiophrys discrimination were used to complete these values. Discussion 1. Habitat a) Breeding Habitat In general, populations of B. a. americanus seemed to be larger and males more clustered. In contrast, it was common to find B. a. hemiophrys males more dispersed, often in scattered small clumps of only two or three calling together. This contrast becomes obvious on examining the sample sizes in collections of the two taxa. Part- ly, this difference in the field is due to the fact that ponds are often larger and deeper in boreal forest and relatively smaller and shallower in aspen parkland and grassland. On the subjective basis of field experience throughout the range of both species, the impression exists that toads in the eastern part of northern North America gather in more extensive and larger breeding aggregations than they do in the central region. b) Summer Habitat Breckenridge and Tester (1961) reported from in- tensive studies of B. a. hemiophrys in Minnesota that it was found most often in or near the margins of prairie ponds. Field observations of B. a. hemiophrys over its Canadian range do not strongly support this as typical of B. a. hemiophrys populations in general. Samples of B. a. hemiophrys (such as the one through the Spruce Woods Forest Reserve in July 1960) have been taken of many individuals foraging well away from water. Surveys made in the summers of 1959-1967 along the margins of lakes, ponds, sloughs, and potholes of all types throughout the range of B. a. hemiophrys seldom revealed large numbers of post-breeding or _post- metamorphosing B. a. hemiophrys at such sites. Exceptions are the Wascana marsh at Regina, Saskatchewan, and the Delta marsh in Manitoba, where large summer series of all sizes of foraging B. a. hemiophrys were taken. Single adult B. a. hemiophrys were occasionally present at pond margins but usually occur there at the end of the breeding season. However, a large series of B. a. americanus of all sizes was taken along a lake margin in eastern Manitoba in August, and occa- sional adult B. a. americanus were found at pond margins near the end of breeding activity. If there is atendency for B. a. hemiophrys to be restricted x7 to pond margins more often than B. a. americanus it is not clearly indicated by present field data. 2. Morphology a) Non-quantitative Live Bufo a. americanus and B. a. hemiophrys are readily distinguished in the field, but their dif- ferences are hard to quantify. Colour charac- teristics, not treated quantitatively, are helpful in this respect. B. a. americanus is usually more brightly coloured, with backgrounds of red, brown, yellow or green. B. a. hemiophrys is a more drab toad, usually greyish or dull brownish. A rusty-coloured phase occurs (Cook 1964c) but it is more subdued than the red of B. a. ameri- canus. Other good field characters, such as distribution of warts, spot size and cranial crest features are treated in detail in the morphological analysis. The difference between the calls of breeding males makes the two taxa easily distin- guishable to the ear. In field sampling across the east-west transects of the interbreeding zone the variety in mor- phology and vocalization gives the impression of sleight-of-hand by a master conjurer. In eastern Manitoba one can collect what appears to the eye and to the ear as ‘‘good’’ B. a. americanus, as recognizable as any from Prince Edward Island to western Ontario. If one misses the intermediate zone, then in the Aspen Parkland east of the Red River one can sample ‘‘good’’ B. a. hemiophrys, as recognizable as any from central Manitoba to western Alberta. When the gap between the two is sampled, a point of change between breeding sites is subtle and almost imperceptible. b) Quantitative A major problem in any morphological com- parison is finding an objective and practical method to evaluate and contrast variations. The discriminant function analysis used here has pro- duced a high-resolution, repeatable analysis of morphological variation over the range of the parent taxa and through the transition zone be- tween them. The reference samples represent a good geographic dispersal for B. a. hemiophrys. Although they are relatively weak in material from North Dakota and the northern portions of Alber- ta and Saskatchewan, all major habitat types which its range covers are represented. The Laramie Valley (Wyoming) disjunct population is also included. The sample largely avoids the southern periphery of the continuous range of B. a. hemiophrys and possible influence from the presumably allopatric B. w. woodhousei. The sample of B. a. americanus is drawn entirely from the northern portions of its range, largely to avoid its extensive area of geographic sympatry over the eastern United States with B. w. fowleri. Only two localities, Rondeau and Point Pelee parks in southern Ontario, are within that contact area. This restriction on the geographic area of samples used means that the discriminant analysis does not represent total B. a. americanus variation. How- ever, the samples do cover three major habitat areas occupied by B. a. americanus in the northern part of its range. Western Ontario is the least well represented geographic area. Encompassed by the samples are areas supposedly occupied by the northern race B. a. copei (Logier and Toner 1961; Ashton, Guttman and Buckley 1973; Conant 1975) and the transition between it and B. a. americanus. In general, the variation in population mean discriminant values within taxa does not markedly follow geographic clines or forest zones though some trends are pointed out in the Results section. However, the wide area of low western scores for B. a. americanus and narrow area of high eastern scores for B. a. hemiophrys may be of some im- portance (see below). Most previous comparisons of B. a. americanus and B. a. hemiophrys (Breckenridge 1944; Blair 1957a; Underhill 1961) have agreed that these taxa are as morphologically distinct as good species, but sample size, geographic representation and number of characters have not been very extensive. A post-orbital score was used by Henrich (1968) as part of his basis for separating the two taxa. It is clearly shown here from the much more ex- tensive geographic coverage of B. a. hemiophrys that post-orbital crests may be present on in- dividuals well within B. a. hemiophrys range and neither their presence nor that of the intermediate conditions clearly establish an individual as intermediate. The ventral scoring used here unfortunately en- compasses only part of the ventral pattern varia- tion exhibited in the two taxa, since it deals solely with the area covered by pigment. As Henrich (1968) pointed out, there is a tendency for B. a. 58 hemiophrys to have lighter ventral markings, and he used a scoring system which attempted to evaluate this difference. However, there is a large subjective element in any attempt to score relative intensity of markings, and this is especially true of trying to assign scores to intermediate specimens. Furthermore, preservation may affect collections differently and there is a danger of scoring preservation differences rather than real phenotypic ones. Similarly, the sometimes obvious difference between individuals in the relative size of markings, spots or reticulations, has not been included in this assessment because of difficulties in obtaining objectivity. Differences are most ob- vious in the northern populations of B. a. americanus and this character does not seem to have a prime importance in the general B. a. hemiophrys-americanus comparison. It is notable in the ventral score results that northern B. a. americanus populations score high, southern score low. The tendency of Prince Ed- ward Island toads to have a high ventral score has been noted previously by Cook (1967). The sug- gestion of Bleakney (1952) that there is a dif- ference between Annapolis Valley populations and those outside this valley within Nova Scotia is not evident in comparing Wolfville (valley) and Mus- quodoboit (Halifax County) samples here. There is great variation in mean snout-vent length in both taxa. The largest population mean lengths (71.8 for B. a. americanus and 70.8 for B. a. hemiophrys) are surprisingly close, although the smallest means (53.9 and 45.8 respectively) are more disparate. When individual maxima and minima are examined, 88.4 vs. 79.1 and 42.1 vs. 41.3 for B. a. americanus vs. B. a. hemiophrys respectively, the agreement is closest in the smallest size range. The discriminant analysis has appar- ently responded to this variation by giving little weight to size in separating the taxa. The poten- tial minimum size at maturity, and the maximum size attainable are, evidently, not that different between the two taxa. Much remains to be understood about geographic variation in size in toads, and the influence of local ecologic condi- tions and yearly variations in rainfall and temperature and therefore on the length of the ef- fective growing period. Perhaps abundance of food and the size at transformation and the total hours per year of optimum foraging weather (in terms of temperature and moisture) have an im- portant bearing on the question. Clear north-south or east-west gradations in either taxa are not ap- parent in the present samples, although there is a generalized trend for B. a. hemiophrys to be smaller in the east, and B. a. americanus to be large in the south and smaller in the north. The relative tibia length has been thought to decrease toward the north in some anurans (Schmidt 1938; Bleakney 1974). This is not clear- cut in B. a. hemiophrys or B. a. americanus populations sampled here, although there may be a trend in this direction in B. a. americanus. Jameson ef al. (1973) have suggested for the Pacific Tree Frog (Hyla regilla) that various pro- portions are correlated with local climatic conditions. | The relative length of the spot (spot length divided by snout-vent length) is much more variable in B. a. americanus than in B. a. hemio- phrys. Western and southern B. a. americanus populations have a smaller mean spot size than B. a. hemiophrys, though the entire B. a. americanus range encompasses the ‘variation in B. a. hemiophrys. It is also noteworthy that there is a difference in dorsal spotting between northern B. a. americanus, where large (and often irregular) spots are present, and southern (and western) populations where smaller spots are more typical. The fact that western B. a. americanus are small spotted argues against the large spots in northern B. a. americanus being derived from contact with B. a. hemiophrys. The separate discrimination for reference samples of females, though based on much smaller numbers (59 B. a. americanus and 118 B. a. hemiophrys) than for males, also separated the two groups completely. The differences in weight- ing may be due to both the smaller sample and the very uneven geographic distribution of female samples (see Appendix I). The Mahalanobis generalized distance between B. a. americanus and B. a. hemiophrys is 104.66 in the female analysis compared with a mean of 51.35 in the male discrimination, possibly indicating that larger body size accentuates the differences in measurements. The discriminant scores for males from southeastern Manitoba clearly define an obvious zone of populations which are intermediate in mean value and have a wider dispersal of scores (as shown by the high variance values) between reference sample populations of B. a. americanus and B. a. hemiophrys. This zone is only a few kilometres wide, and its narrowness most sharply demonstrated on the Trans-Canada transect where 59 intensive sampling indicates it is most pronounced over 10 miles [16 km] or less in width. The nar- rowness of the zone is also apparent in the other five transects. The histograms show no clear in- dication of modality in the distribution of scores which would have been indicated in hybrids mixed with ‘‘pure’’ examples of either or both parental types. Instead, the distribution seems best inter- preted as complete interbreeding and_ back- crossing with no segregation into B. a. americanus or B. a. hemiophrys types within the interbreeding zones. The lowered scores east and west of the most obvious interbreeding zone could be interpreted as due to convergence in each taxon toward characters of the other as the intermediate zone is approached. This interpretation would require selection to be acting on the natural variation within each of the B. a. hemiophrys and B. a. americanus stocks and would ignore the likelihood of gene flow from the intermediate zone. Discrimi- nant scores of B. a. hemiophrys do not tend toward those for B. a. americanus over the north- ern part of the former’s range where it occurs in boreal coniferous forest, nor do scores of north- eastern B. a. americanus tend toward B. a. hemiophrys where B. a. americanus also occurs in boreal coniferous forest. Because the lower scores grade relatively smoothly toward the area of contact, it is evident they reflect the direct ef- fects of introgression from the contact area. This distribution of lowered mean scores only as far as Delta in central Manitoba, but all the way to Whitetop Creek, northeast of Moosonee in cen- tral Ontario, suggests that introgression is more successful into B. a. americanus than into B. a. hemiophrys populations. That this is really occur- ring is further indicated by the observation of Logier (1928) of a specimen with ‘‘temporal crests’’ (cranial crests) resembling B. a. hemio- Dhrys in a sample of toads from Lake Nipigon (geographically between Oxdrift and Whitetop Creek samples analysed here). Characters analysed separately or in ratios also show the same pattern of abrupt change over a narrow zone that is evident in the multivariate analysis. Cranial crest score, cranial crest posterior width/head width, and ‘“‘spade’’ width/tarsus length show the abrupt change in their means across this zone, as does the spot length/snout- vent length ratio. The latter value is different be- tween western B. a. americanus and all B. a. hemiophrys. It did not weight heavily in the discrimination, but this may have been because of its broad variation over the whole of B. a. ameri- canus distribution. The scores for females from southeastern Manitoba show a similar pattern to that obtained with the scores for males though the narrowness of the transition zone between the two taxa is not as well defined by the former. This may be partly due to the meagre samples of females available, as the largest samples show the greatest agreement with the results obtained for males. c) Selective Advantage of Taxonomic Characters B. a. hemiophrys has a major portion of its range in Aspen Parkland and Prairie regions and is therefore subject to more arid conditions and greater environmental fluctuation than B. a. americanus. The latter inhabits the mesic Great Lake-St. Lawrence, Acadian and Eastern Deciduous forests over much of its range. Both occur widely in the Boreal Forest in the northern portions of their ranges, and B. a. americanus con- tinues into the Boreal-Tundra transition. Harper (1956, 1963) has pointed out that the northern limit of each is reached south of the limit of permafrost. The appearance of the cranial crests is the most striking difference between B. a. hemiophrys and B. a. americanus. This difference may be related to selection pressures arising from differences in soil type and aridity in the respective environments of the two forms. Although experimental evidence is lacking, the cranial crests may function as the protection for the eyes, since the latter can be withdrawn into the head below the level of the crests. The solid or grooved boss formed by the crests of B. a. hemiophrys may be of advantage in emerging from the heavy prairie soils found over much of its range. Similar thickening, or thickening and filling in of the crests is found in the closely related B. w. woodhousei, the more distantly related B. cognatus, and in the Plains Spadefoot, Scaphiopus bombifrons, which belongs to a separate family. All occur in prairie habitats over much or all of their ranges. In con- trast, the eastern forest-dwelling B. a. americanus has well-separated, relatively narrow crests (though they may be pronounced posteriorly) as does the eastern B. w. fowleri. The Eastern Spadefoot, Scaphiopus holbrooki, which typically occurs on light sandy soils, lacks a boss. The toad of the northern forests of western North America, B. boreas, either lacks cranial crests, or has only weak, low traces of them. 60 The relatively shorter tarsus of B. a. hemiophrys would result in shorter muscles in this portion of the leg, which may provide additional strength at the expense of some agility compared with B. a. americanus. This could be an advantage in dig- ging in heavier prairie soils and for digging to greater depths to avoid aridity and frost. The relatively large spade in B. a. hemiophrys would also be an adaptation for more efficient digging in more difficult conditions. B. cognatus has a large spade as well, whereas this structure is relatively narrower in B. a. americanus and much narrower in B. boreas. The parallel reduction of the inner metatarsal tubercle that seems coincident with a large spade is noteworthy as the spadefoots (Scaphiopus), which are primarily arid-adapted, lack this tubercle entirely. Tihen (1962a) postulated that the development of metatarsal spade and heavier limbs is associated with burrowing ability as an adaptation to arid conditions in the ‘‘americanus’’ group in which he includes B. cognatus. Martin (1973) found the tibiofibula and femur of B. a. hemiophrys to be shorter and broader than all other members of the B. americanus group and the humerus shorter than it is in all but B. woodhousei. He suggested these proportions were due to selective pressures for efficient burrowing apparatus ‘‘in the cold environment’’ of this species. Other characters, such as the nostril separation, head width and parotoid gland separation are related to a wider, stockier build in B. a. hemio- Dhrys and B. cognatus compared with the forest- dwelling forms. If B. a. americanus, exclusive of its northern- most populations, is compared with B. a. hemiophrys, the latter has larger dorsal blotches containing smaller and more numerous warts. B. cognatus shows a more extreme trend in this direc- tion. However, northern B. a. americanus popula- tions have large blotches and generally more warts per blotch. These northern toads may be more diurnal because of longer day length and cooler nights and occur in more open habitats than more southerly B. a. americanus. Larger blotches may provide a more contrasted pattern and better camouflage for a toad in open habitats. The more subdued colours in B. a. hemiophrys, where grey coloration is common basic background, is evi- dent also in B. cognatus. Both lack the deep browns or reds that are evident in some B. a. americanus and B. boreas, and reach extreme development in northern populations of the former. Nevo (1973) in a study of colur polymor- phism in cricket frogs (Acris) found the frequency of a grey morph increased, and red and green morphs decreased, along increasingly arid habitat gradients. While there is general agreement that B. a. americanus is a forest toad, some midwestern populations may show prairie adaptations. Smith (1961) designated it as a ‘‘prairie toad’’ in northern Illinois. However, he also expressed reservations about the equivalence of these populations to B. a. americanus Holbrook. Garman (1892) noted that a B. americanus he examined from Illinois had the cranial crests bridged posteriorly but did not indicate where in the State it had been col- lected. Smith (1961) regarded B. a. charlesmithi of southern Illinois as a forest toad. The varia- tion and ecology of toads in this and adjacent states deserves further attention, particularly as the post-glacial prairie peninsula once extended over the northern portion of Illinois. 3. Breeding Call Analysis and Comparisons Blair (1957a) has previously pointed out the distinctions and resemblances between the calls of B. a. americanus and B. a. hemiophrys and this comparison led him to postulate their close rela- tionship. Henrich (1968) did not have data available for his comparison of these taxa. The samples for call analysis were restricted to a much smaller geographic area than those available for morphological analysis. Only those tapes recorded during 1969-1970 had body temperatures of calling toads recorded. Zweifel (1968) has stressed the importance of using actual body temperature in analysis of call parameters in preference to the air or water temperatures in the vicinity of a recorded animal. The body temperature of a toad may be close or identical to either that of the air or water in its vicinity. However, body temperature can also have an intermediate value because of recent movement from one medium to the other, and even adjacent toads may have different body temperatures. The results obtained in this study of the relative effects of temperature and snout-vent length on the call parameters (pulse rate, dominant fre- quency and duration) are in general agreement with those of Zweifel (1968) who compared samples of 50 B. a. americanus from two popula- tions in New Jersey and 63 B. woodhousei fowleri 61 from New Jersey and New York populations. Zweifel found a strong positive correlation be- tween temperature and pulse rate in both species and a negative correlation similar to the one reported here for temperature and call length. He did find temperature and dominant frequency to be correlated for Northville, New Jersey, B. a. americanus (r = + 0.38, P <0.05) and this is quite similar to the B. a. americanus result in this study (r= + 0.416, P<0.01), but his Hayworth, N.J., sample showed a weak negative correlation which was not significant (r= —0.08, P>0.1). Zweifel’s comparisons of body size with pulse rate, call length and dominant frequency for B. a. americanus are similar to the results of this study in that he did not find any significant correlation. His results for B. w. fowleri in comparisons of body size, however, gave a significant correlation (+0.47, P=0.001) with call length in the Hayworth sample and correlations with dominant frequency of —0.49, (P<0.1>0.05) in a Long Island sample and —0.59 (P<0.001) in the one from Hayworth. A comparison of Zweifel’s values for the two most highly correlated variables, pulse rate and temperature, for his two B. a. americanus and two B. w. fowleri samples with samples analysed in this study of B. a. americanus and B. a. hemiophrys shows the similarity of the New Jersey and Manitoba-western Ontario populations of B. a. americanus. As well, it indicates that B. a. hemiophrys occupies an intermediate position be- tween B. a. americanus and B. w. fowleri (Figure 38). The separation of taxa based on call and associated variables is equally as effective as the separation on strictly morphological variation. Intermediate populations show the wide variability which is typical of the morphological analysis. However, in the call discrimination the body temperature of the toad has made a contribution, apparently because of the relatively fewer B. a. hemiophrys recorded at low readings. The comparison of the discriminant scores for call and morphology indicates that call variation does not follow morphological variation within either taxon, but that in the intermediate popula- tions there is a tendency for call type to correlate with morphological type. Within the contact zone B. a. hemiophrys-like toads morphologically tend to have B. a. hemiophrys-like calls and B. a. americanus-like morphologically tend to have B. a. americanus-like calls. The comparison of the =) z fe) 1S) wu ”n \ nN uw 72) = 2 a oO [als fal a oO O pOOOp O > Of OOB8D POOOO O8 OO oO FOWLERI( ZWEIFEL 1968) HEMIOPHRYS INTERMEDIATE AMERICANUS AMERICANUS (ZWEIFEL 1968) Figure 38. Relationship between pulse rate and temperature in three samples of Bufo from this study, and two B. a. americanus and two B. w. fowleri populations from northern New Jersey and Long Island, New York (data from Zweifel 1968: Figure 8). 62 total pooled sample shows the same correlation of call and morphology between taxa (Table 8). 4. Crossbreeding Experiments The most important result of these crosses is that all combinations can produce metamorphosing in- dividuals. This supports the morphological evidence that natural intermediates are freely pro- duced. Unfortunately, the experiments were ter- minated at transformation. However, the success of toads from intermediate localities, as well as those from the most separated localities, indicates that no absolute fertility barriers exist. The dif- ferences in results in some of the replicate crosses and the unequal number of replicates makes any conclusion based on relative success between com- binations unreliable. Previous hybridization experiments summarized in Blair (1972) involving B. a. hemiophrys and B. a. americanus have also demonstrated their abili- ty to produce metamorphosing offspring. Porter (1968) demonstrated fertility between individuals from the Wyoming relict population of B. a. hemiophrys and those from North Dakota and Manitoba. Blair (1972: Appendix H) reported two artificial crosses involving a female B. a. americanus and a male B. a. hemiophrys in which 85.2% and 34.9% of the resulting larvae metamorphosed. Three crosses involving a B. a. hemiophrys female and an B. a. americanus male gave 64.3, 93.9 and 43.5% fertilization of eggs. In two of the crosses 37.8% and 86.9% of the fertilized eggs hatched and 42.0 % and 29.4% of the larvae reached metamorphosis. 5. Taxonomic Conclusions a) Comparisons of B. a. hemiophrys and B. a. americanus with other Bufo in northern North America (i) B. a. hemiophrys-B. boreas: In southern Alberta, B. boreas is restricted to the Rocky Mountains and their foothills, whereas B. a. hemiophrys occurs in Grassland and Aspen Parkland regions to the edge of the foothills. To date, no contact has been found in this region (un- published results of herpetofaunal surveys of Canadian Prairie Provinces). However, north of Edmonton, B. boreas occurs west of the foothills in the Boreal Forest region and has a narrow area of sympatry with B. a. hemiophrys. To the north- 63 west, only B. boreas occurs in Aspen Parkland islands of the Peace River region, an area where, by habitat, B. a. hemiophrys would be expected. Eastern Alberta B. boreas, although larger than B. a. hemiophrys at these localities, are markedly smaller than typical B. boreas from British Co- lumbia (unpublished NMC data). Also, although it has been often asserted that B. boreas lacks a true breeding call (Blair 1972 and elsewhere), populations in eastern Alberta have a distinct call, which has a low dominant frequency and is very slowly pulsed. It has been noted at Banff, High River, and in the overlap zone with B. a. hemiophrys (unpublished NMNS data). Although very distinct from the higher, rapidly pulsed call of B. a. hemiophrys, the frequency of mismatings in mixed choruses throws doubt on its complete effectiveness as an isolating mechanism within a common breeding pond. Different calls may, however, serve to differentially attract females and males when breeding ponds are distinct. Blair (1972: Appendix H) reported two artificial crosses between female B. a. hemiophrys and male B. boreas in which 41.7% and 79.7% of the eggs were fertilized and 77.4% and 3.2% of the larvae metamorphosed. Only one cross involving a female B. boreas and a male B. a. hemiophrys was reported and this resulted in 98.2% fertilization and 81.8% hatching but none of the resulting lar- vae reached metamorphosis. No naturally occur- ring hybrids have been previously reported be- tween these taxa. The B. boreas-B. a. hemiophrys overlap without loss of distinctness and only rare survival of natural hybrids provides an interesting contrast to the B. a. americanus-hemiophrys interactions. However, the narrowness of the overlap, the ap- parent similarity in breeding time and site, and the closeness in body size may be indications that these two forms have not diverged enough to partition an area between them. (ii) B. a. hemiophrys-B. cognatus: B. a. hemiophrys and B. cognatus are sympatric in the shortgrass prairie or southern Alberta and Sas- katchewan and the adjacent Dakotas and western Minnesota. B. cognatus ranges south into Mex- ico (Stebbins 1966), is much more grassland adapted than B. a. hemiophrys and their sympatric area, which is much broader than that of B. a. hemiophrys and B. boreas, represents the area where B. cognatus reaches its northern limit and B. a. hemiophrys, with the exception of the Wyoming relict, reaches its southern limit. In Min- nesota, a natural hybrid between B. cognatus and B. a. hemiophrys has been reported (Brown and Ewert 1971). B. cognatus has a very long distinctive call resembling the sound of a pneumatic hammer (Co- nant 1975). This call has tremendous carrying capacity and is strikingly distinctive from calls of the B. americanus group. The vocalization of males in this chorus could be heard at least 0.5 mile [0.8 km] over the prairie from the road where we had stopped our vehicle. B. a. hemiophrys was not heard in this din until we were much closer to the pond. Blair (1972b: Appendix H) has recorded one ar- tificial cross of a female B. a. hemiophrys and a male B. cognatus where 68.0% of the eggs were fertilized, 73.5% of those fertilized hatched, but only 3.5% of the larvae metamorphosed. In ad- dition, in one cross of a B. cognatus female and a B. a. hemiophrys male 33.3% of the eggs were fertilized but none hatched, all stopping in gastrula or neurula stage. Gi) B. a. hemiophrys-B. w. woodhousei and B. a. americanus-B. w. fowleri: The subspecies B. w. woodhousei is not known to extend into Canada and in the northern United States. Its range as cur- rently understood is allopatric to that of B. a. hemiophrys. Blair (1972: Appendix H) reported two crosses using a female B. a. hemiophrys and a male B. w. woodhousei. One of these resulted in 97.0% fer- tility, with 83.6% of the fertilized eggs hatching and 26.6% of the larvae reaching metamorphosis. For the other the only data given are that seven larvae were obtained and 85.6% (or six of them) metamorphosed. In two crosses between a B. woodhousei female and a B. a. hemiophrys male given in the same report, 100% were fertilized and 24.4% of 600 larvae metamorphosed in the first and 31.2% of 279 larvae metamorphosed in the second. These values for hatching success are only marginally lower than those obtained with crosses between B. a. hemiophrys and B. a. americanus. The subspecies B. w. woodhousei has promi- nent, usually somewhat parallel cranial crests, few warts per dorsal blotch and generally has an unspotted venter. Blair (1972) mentions that Great Basin populations are smaller and may have a cranial boss, but there is no published comparison of this variant with prominently bossed B. a. hemiophrys. B. w. woodhousei occurs primarily 64 in the great plains of the central United States, often in river valleys, and is sympatric over much of its range with B. cognatus, which tends to oc- cupy the upland prairies (Timken and Dunlap 1965). The other northern subspecies of B. woodhousei, B. w. fowleri, is eastern in distribu- tion and largely sympatric with B. a. americanus. It is smaller than B. w. woodhousei and generally has more warts per dorsal blotch. Its intergrada- tion with B. w. woodhousei has been documented by Mecham (1962). Underhill (1961b) has com- pared morphological variation between B. a. hemiophrys and B. w. woodhousei in South Dakota but his sample of the former was probably contaminated by introgression from B. a. americanus (cf. localities given by Underhill 1961 and Henrich 1968). The call of B. w. fowleri has a faster pulse rate than that of B. a. hemiophrys (Figure 38). It is interesting to note that, while B. w. fowleri is in- termediate between B. a. hemiophrys and B. a. americanus morphologically in the characters studied here, B. a. hemiophrys is the intermediate with respect to this call variable. The B. a. hemiophrys samples geographically closest to the northern limit of B. w. woodhousei (those from the prairie grassland of southern Alberta and southwestern Saskatchewan) have low mean discriminant scores (— 30.8 to — 26.9). These give no indication of introgression from interbreeding with B. w. woodhousei further south as, if this was affecting southern B. a. hemiophrys scores it would presumably raise them. Another indication that interbreeding of B. w. woodhousei and B. a. hemiophrys is not widespread is the survival of the Wyoming disjunct of B. a. hemiophrys within B. w. woodhousei range. However, the narrowness of the contact zones between other Bufo taxa where complete interbreeding can occur leads one to be cautious in drawing any but tentative con- clusions until adjacent populations can be analys- ed. The possibility exists that introgression from B. w. woodhousei had some effect on the scores obtained by Henrich (1968) for his 160-mile [257.6 km] intergradation zone between B. a. hemiophrys and B. a. americanus in South Dakota. Much of this zone is a southern extension of B. a. hemiophrys which is not only adjacent to B. a. americanus on its east but also to B. w. woodhousei on its west. Comparison with B. w. woodhousei was not made by Henrich. Bufo hemiophrys was considered a subspecies of B. woodhousei by Schmidt (1953) but Blair (1957a) presented convincing evidence for its closer rela- tionship to B. americanus. Similarly, the mean discriminant value for B. a. americanus samples adjacent to the range of B. w. fowleri (+ 33.8 at Rondeau and + 35.3 at Point Pelee) in high, perhaps even indicating divergence (character displacement of Brown and Wilson 1956) between these forms in adjacent localities. Elsewhere in a wide area of geographic sympatry the two taxa are known to hybridize (e.g. Zweifel 1958), and selection against hybrids may be an im- portant factor here. b) Status of B. americanus copei and B. hemiophrys baxteri (i) B. a. copei: Northeastern B. americanus Populations: The validity of the northeastern Bufo americanus copei (Yarrow and Henshaw 1878) is open to question. Unmistakably, there are popula- tions of brightly coloured B. americanus with con- trasted patterns accentuating reds and oranges in the northern portion of its range. These popula- tions have also been characterized as having heavi- ly pigmented venters and shorter hind legs. Such characters formed the basis for the initial recogni- tion as B. copei as a distinctive form, and for Gaige (1932) to have resurrected it from synonomy as B. a. copei. It was first described from James Bay material, but subsequently Trapido and Clausen (1938), Grant (1941), Vladykov (1941), Netting and Goin (1946), Backus (1954) and Harper (1956) extended its range across northern Quebec. Schueler (1973) reaffirmed the distinctive coloration of James Bay toads. Bleakney (1952) applied the name to Nova Scotia populations out- side the Annapolis Valley which had shorter tibias and more heavily pigmented venters. Subse- quently, however, Bleakney (1958) did not recognize subspecies within B. americanus in eastern Canada. Most recently, Ashton, Guttman and Buckley (1973) extended the range of B. a. copei south over much of central Ontario and Quebec. However, B. a. copei characters are not pronounced in all northern individuals (Cook 1964, 1968). Earlier Logier (1952) and Logier and Toner (1955, 1961) presented a discussion of the validity of the race in which they pointed out the intermediate nature of supposedly diagnostic characters across a wide geographic area. Of all these papers, only Logier (1952) presented any statistical comparison of B. a. copei and B. a. americanus populations, but only relative leg 65 length was treated. In the present study neither tibia nor tarsus length appear to be useful in defin- ing a distinct northern race within B. a. americanus. Two measurements of variation in the present analysis which do have bearing on the status of B. a. copei are the length of the dorsal spot and the ventral pigmentation score. Particularly high mean values for these characters are present in northern population samples: Whitetop, Moosonee, Lake Attila, Sept Iles, Mile 134 [Kilometre 215.7], and Routhierville. To the south there is a wide zone of intermediates until the “*typical’’ (southern) B. a. americanus with smaller dorsal spots and a less heavily pigmented venter is found. The variation in coloration seems to parallel these characters with the most brightly col- oured toads in the north. The wide dorsal stripe and more contrasted pattern mentioned by most observers recognizing B. a. copei are also evident in the material examined here and specimens with large dorsal blotches generally have these features as well. The dorsal ‘‘stripe’’ cited by previous authors seems to be an area largely defined by the dorsal blotches, and may therefore be highly ir- regular in width, varying with evenness of blotch edges and difficult to measure objectively. A ‘*trace’’ mid-dorsal stripe, which is very narrow with parallel edges, can also be seen on many specimens and is somewhat paler than the broad stripe. This narrow stripe is widespread in B. americanus group toads. Northern B. americanus are apparently often strongly diurnal in habit (personal communica- tions: Ross MacCulloch, northern Quebec; F.W. Schueler, James Bay) in contrast to the generally nocturnal or crepuscular behaviour, especially of adults, further south (FitzGerald and Bider 1974: 65 miles [104.7 km] NW of Montréal, Quebec). Diurnal activity is to be expected in the north because of long day length and cool evening temperatures during the short summer. The bright colours and highly contrasted pattern of northern populations may be of selective advantage during diurnal activity because of the disruptive visual ef- fect of such contrasts. Because of the wide area which is transitional from north to south, there seems little justifica- tion in distinguishing northern populations as a distinct subspecies. As Logier and Toner (1955, 1961) have pointed out, the intergrade zone would be wider than the range of the northern race. Ashton, Guttman and Buckley (1973) attempted to resolve this difficulty by redefining B. a. copei to include most of these intermediate populations but failed to provide evidence that this enlarged concept of B. a. copei could be adequately distinguished from southern B. a. americanus. Guttman (1975) failed to find distinctive isozyme distributions between B. a. copei and B. a. americanus. The reduction in blotch size and ven- tral pigmentation in B. americanus is toward its southern area of sympatry with B. w. fowleri. It is suggested here that the transition between ‘*copei’’ and southern B. americanus is due in part to different selective pressures on extreme north- ern and southern populations and a cline through the intervening area. (ii) B. h. baxteri: the Wyoming Disjunct of B. a. hemiophrys: B. h. baxteri was named by Porter (1968) from a comparison of 34 Wyoming specimens and an equal number from a composite sample drawn from North and South Dakota and Manitoba. The differences used to define the subspecies were mainly statistical comparisons of selected morphological characters. He concluded that Wyoming toads tended to be smaller, with shorter radio-ulna and tibio-fibula measurements, narrower heads, a more prominent boss and a narrower mid-dorsal stripe. Ulna length, boss height and mid-dorsal stripe were ex- cluded from the present study. However, measurements of snout-vent length, tibia length and head width (remeasured on most of Porter’s sample) do not distinguish Wyoming populations when considered against the variation in popula- tions from throughout the range B. a. hemiophrys. The relatively small differences found by Porter in the characters not measured here would prob- ably not validate the distinctiveness of the Wyo- ming B. a. hemiophrys. The discriminant score mean for the Wyoming population is well within the range of score means over the rest of the geographic range of B. a. hemiophrys. Criteria for subspecies vary widely between workers, but commonly a significant proportion of individuals within a geographic area have to be distinguishable from all other individuals in other areas to warrant taxonomic recognition (the so- called 75% rule). Smith (1974) has argued that an ‘‘abrupt geographical step’’ in the variation of at least one character should be present to warrant subspecies recognition. B. h. baxteri does not ap- pear to meet either criterion for a subspecies. A 66 disjunct distribution is not generally recognized as sufficient cause for erecting of a taxon unless distinguishing morphological or other characters can be clearly defined (Mayr 1963). The over- emphasis of the significance of its geographic isolation even led Packard (1972) to recommend the recognition of Bufo baxteri as a full species. However, Porter (1968) presented evidence that Wyoming toads were successful in crossing with Manitoba and Dakota toads under laboratory con- ditions. There is no reason to suppose they would not interbreed freely in nature if re-expansion of range brought the Wyoming population into con- tact with any other B. a. hemiophrys populations. Porter (1968) found only slight difference in the call of the Wyoming population. As he did not record toad body temperatures exact comparison of his data with the variation presented here could not be attempted, but I do not believe that the dif- ferences appear significant. c. The Status of B. a. americanus and B. a. hemiophrys The contrast between results in the analysis of southeastern Manitoba samples compared with the reference samples from throughout the range of B. a. hemiophrys and northern B. a. americanus is evident. Clearly the distinction between the mor- phologically ‘‘good species’’, ‘‘B. americanus’’ and ‘‘B. hemiophrys’’, disintegrates biologically across this zone, and extensive interbreeding and backcrossing are freely occurring in the apparent absence of strong isolating mechanisms. The nar- rowness of the contact zone between the taxa is emphasized by both morphological and call analysis. Its limited extent may be a function of the relative abruptness of the vegetation ecotone where it occurs. It may also be affected by the distance moved by individual toads in their lifetime. This situation may be designated allopatric hybridization in the sense of Woodruff (1973). Both parental forms are distinct on either side of a zone populated by intermediate animals but pure individuals of either parent are not clearly distinguishable within the intermediate zone. This term contrasts with Woodruff’s parapatric hybridization, where both parental forms are iden- tifiable in an intermediate zone containing hybrids, and sympatric hybridization, where the ranges of the parental forms may extensively overlap and hybrids are produced in varying numbers. These interactions represent various gradations between the relatively free and com- plete gene flow generally assumed to occur within a species and the absolute reproductive barriers (no gene exchange) between indisputably good species. The various natural interactions which demonstrate some, but incomplete, isolation be- tween populations have been designated as semi- species, allospecies or incipient species as exten- sively reviewed by Amadon and Short (1976). Mayr (1963, 1969, 1970) clearly regards cases where extensive interbreeding occurs between two essentially allopatric populations as evidence of their lack of reproductive isolation and concludes that such populations belong to the same species. Rising (1970) reviewed some of the inconsistent treatment of a selection of such taxa by many workers. Bigelow (1965) argued that reproductive isolation rather than interbreeding is the essential part of Mayr’s definition and that the former could exist and be maintained even if interbreeding produces a range of recombinations of the paren- tal genotypes in a hybrid zone. Although such forms may have no barriers to interbreeding where they meet, he reasoned, strong selection on either side of the hybrid zone which eliminates variants carrying genes of the other form would produce effective isolation. Short (1969) suggested that Bigelow ignored the differential action of natural selection on characteristics. One or a few genes from one differentiated population may be incor- porated into another due to interbreeding and, as selection is unlikely to be so severe that some in- trogression is not occurring, complete reproduc- tive isolation is not likely to be maintained by selection where there are no barriers to free inter- breeding. Only the sympatric occurrence of the two forms can be taken as an indication of reproductive barriers. Mayr (1963) largely dis- counted introgression as of any consequence be- tween species in animals, as he regarded species’ gene pools to be internally cohesive and coadapted that they are largely resistant to incorporating genes from other gene pools. Sibley (1969) has provided a framework for tax- onomic judgement of the varying grades of inter- breeding situations. Evidence, or lack or it, for some degree of reduction in interbreeding as shown by at least some degree of sympatry is judged to be of paramount importance. The natural occurrence of some hybrid individuals or some geographic areas where complete inter- 67 breeding occurs (including isolated instances of local hybrid swarms) is weighed against the geographic extent or numerical preponderence of *‘nure’’ forms or populations in contact. If con- tact without extensive interbreeding predominates, the taxa have species status. But, if in the majori- ty of contacts, interbreeding between distinct taxa produces a population in which all or nearly all individuals have some degree of intermediate characteristics the taxa are regarded as subspecies. Their gene pools are obviously not isolated. Schueler and Rising (1976) provided a thoughtful analysis of cases where ‘‘hybridiza- tion’’ could be recognized, stressing the impor- tance of mean, variance and modality of discrimi- nant analysis scores. Although they would regard results such as obtained here as indicating hybridization, it is unclear if they would consider species status a mandatory consequence. In the present case, interbreeding occurs be- tween B. a. americanus and B. a. hemiophrys throughout their contact zone in eastern Manitoba. Henrich (1968) has clearly demonstrated the same situation at the opposite end of their contact, in South Dakota. Although Henrich’s zone as presented is apparently much wider, his measure (a coefficient of similarity) may not be as precise for delimiting the contact zone as the discriminant function, and therefore the results of the two studies are not exactly com- parable. His sampling area ran in a NW-SE direc- tion, and may be cutting the contact area on a diagonal, rather than at right angles, thus not evaluating the true width of contact. A similar ef- fect is noticeable in the ‘‘southern’’ transect presented here. It has been suggested earlier that analysis of the South Dakota contact should also consider possible introgression from B. w. woodhousei on the west, which could introduce additional variability, and perhaps broaden the ap- parent contact area. Minnesota, in contrast to the northern and southern contact in Manitoba and South Dakota, respectively, has not received comparable analysis. That the contact zone there is narrow is apparent from statements by Breckenridge (1944), Blair (1957a), and Tester, Parker and Siniff (1965), who had no difficulty in assigning collections to one taxon or the other, and doubted any range overlap existed. Henrich included a few Minnesota specimens in his analysis and found evidence of intermediacy of characters. Guttman (1969) found some transferrins with identical mobilities in Min- nesota B. a. hemiophrys and B. a. americanus. This evidence indicates a continuous inter- breeding zone between the southern end of Lake Manitoba through Minnesota to northern South Dakota wherever B. a. hemiophrys has come in contact with B. a. americanus. The contact is cen- tred on the ecotone between Great Lakes and Deciduous Forest with Parkland and Prairie (this study; Henrich 1968). No evidence exists of primary isolating mechanisms (inability to physically mate or ab- solute sterility of hybrids) or strongly developed secondary isolating mechanisms (ecological or temporal) (Mayr 1963). The difference in calls may, however, be a weakly effective behavioural isolating mechanism. Licht (1976) in a study of B. a. americanus has suggested that female toads select mates, and has presented some observational evidence. He hypothesizes that call variables of vocal males allow females to select a male within a narrow size range. The evidence for call variables following precisely enough with male snout-vent length within a single taxon to allow such discrimination is weak (this study; Zweifel 1968). If the choruses in the intermediate zone consisted of clearly definable parental forms and hybrids, such discrimination by females might be reasonably postulated. However this is not the case. The ma- jority of toads in the contact zone are intermediate in both call and morphology, and no clear assort- ment into parental types and hybrids is evident. Mate selection by females, if it occurs here, may be effective at the edges of the contact zone in allowing a female to select a ‘‘pure’’ male, or chorus, instead of a hybrid individual or chorus of hybrids. This may increase in effectiveness with distance to the east and west of the zone. It is in- teresting to note in this regard that the B. a. hemiophrys males west of the contact zone are the smallest that occur (both in range and mean). Some character displacement (Brown and Wilson 1956) in size may presently be occurring in B. a. hemiophrys. This may account, in part, for the reduced extent of apparent introgression of B. a. americanus genes into B. a. hemiophrys populations. It seems reasonable to assume that if B. a. hemiophrys characters are adaptations for relatively more arid environments, and if selection is most severe on toads in arid habitats because of recurrent drought then, ‘‘mesic-adapted’’ genes 68 may be less able to penetrate westward than ‘‘arid- adapted’’ genes can penetrate eastward. The geographic distribution of discriminate scores sug- gests that this could be true. Higher variances in discriminate scores on the eastern side of the con- tact zone, and the possible character displacement in size of B. a. hemiophrys to the west support the possibility of more restricted gene flow to the west than to the east of the contact zone. Tester, Parker and Siniff (1965) concluded, from behavioural studies, that B. a. americanus is less well adapted to avoid heat stress in open areas than B. a. hemiophrys. This could impose further restriction on the westward flow of B. a. americanus genes. In summation, although B. a. americanus and B. a. hemiophrys are clearly as distinct as good species in morphology and call where allopatric, the lack of any recognizable sympatry between pure forms of each indicates a lack of isolation mechanisms where they contact each other. They cannot be considered as having isolated gene pools and therefore should be treated as subspecies (Sibley 1961; Short 1969; Mayr 1963; Woodruff 1973; Amadon and Short 1976). The morphological distinctness of B. a. americanus and B. a. hemiophrys and the high variability of hybrid populations indicates that their respective gene pools have incompatabilities and are perhaps somewhat differentiated. Amadon and Short (1976) have used the term megasubspecies for well-marked subspecies known, or judged to be, approaching the status of a species. In trinomials designating a megasub- species the species name is placed in parentheses, and the taxa here should be written: Bufo (americanus) americanus Bufo (americanus) hemiophrys This notation complements Amadon’s (1966, 1968) suggestion that allospecies should be listed with the superspecies of which they are a member placed in brackets. The distinction between allospecies and megasubspecies is developed from the discussion given earlier by Short (1969). When the designated forms contact a closely related form a predominance of allopatric hybridization in the sense of Woodruff (1973) indicates megasub- species and a predominance of parapatric or sym- patric hybridization indicates allospecies. The utili- ty of this notation will be mainly in zoogeographic studies to clarify the relationships of taxa where comparison of faunas of distinct areas is made. For most purposes the brackets may be omitted. 6. Zoogeographic Significance of the Eastern Manitoba Transition Zone The significance of location of the B. a. americanus-hemiophrys transition and its origin can be related to the general herpetofauna distribution patterns of northern North America. Eastern Manitoba is the centre of a transition between eastern and central herpetofaunas (Cook 1974). The B. a. americanus-hemiophrys contact line defined here also is the approximate western limit in southeastern Manitoba for at least two eastern species, the Blue-spotted Salamander, Am- bystoma laterale, and the Spring Peeper, Hyla crucifer. It is the western limit for at least one cen- tral species, the Plains Garter Snake, Thamnophis radix. In addition, the wide intergrade zone be- tween the central Red-sided Garter Snake, Tham- nophis sirtalis parietalis, and the Eastern Garter Snake, 7. s. sirtalis, which extends from extreme western Ontario to central, Manitoba, is approx- imately centred on this line (Cook, unpublished data) and part of the overlap between the Gray Treefrogs Hyla chrysoscelis (western) and Hyla versicolor (eastern) in Manitoba occurs here (Stewart and Cook 1970, and unpublished data). This contact zone may be regarded as a zoogeo- graphic suture zone in the sense of Remington (1968), a region where two faunas thought to be previously separated have re-established contact resulting in a variety of interactions. Because the northern faunas are poor in species diversity, the number of interactions are few. As pointed out above, they vary from the zone of sympatry of ap- parently good species in Hyla versicolor and H. chrysoscelis to the wide area of subspecies inter- gradation between 7. s. parietalis and T. s. sir- talis. The B. a. americanus-hemiophrys inter- action is intermediate between these extremes, showing a sharply demarcated contact with limited introgression. Western limits for widely distributed eastern amphibians and reptiles correspond to gradually increasing aridity from east to west (Cook 1975; see also Figure 77, p. 295, in Bryson and Wendland 1967), as do, in reverse, eastern limits for many central forms. A marked change in relative aridity in the Manitoba lowlands is in- dicated by the sharp break between largely con- iferous forest and aspen parkland. The B. a. americanus-hemiophrys peak interbreeding seems centred in this area. The central herpetofauna as defined by Cook 69 (1974) is bounded by this eastern Manitoba suture- zone in the east and by the foothills of the Rocky Mountains in the west. The majority of amphi- bian and reptile species which range north into the Boreal Forest areas in each of the eastern, central, and western regions between these boundaries are specifically or subspecifically unique to each region. The transcontinental Boreal Herpeto- faunal Zone postulated by Savage (1960) extend- ing from Alaska to the Atlantic coast, seems based primarily on one species, the Wood Frog, Rana sylvatica, which is the sole taxon which ranges widely across all three regions. However, the Wood Frog is not restricted to Savage’s supposed Boreal Zone but occurs far south of it in the east continuously and in the west as disjunct popula- tions. In the central region, all Boreal species (Rana sylvatica; the Leopard Frog, Rana pipiens; the Boreal Chorus Frog, Pseudacris triseriata maculata; B. (a.) hemiophrys and T. s. parietalis) also occur throughout the adjacent Aspen Parkland. In the east, three wide-ranging Boreal species, Ambystoma laterale, Hyla crucifer, and the Mink Frog, Rana septentrionalis, reach their western limit in southeastern Manitoba at or near the Aspen Parkland boundary. None invades the Aspen Parkland. Two others, B. (a.) americanus and T. s. sirtalis, are replaced over this boundary by central subspecies. None of these five eastern forms occurs in the central Boreal Forest although suitable habitat appears to exist for them there. However, all central Boreal Forest species either show some range extension or genetic influence to the east. In R. pipiens no differences between central and eastern northern populations have been defined. In R. sylvatica a striped morph, common in central populations, shows a decreas- ing abundance from west to east in northern Canada. The number of individuals possessing a stripe shows its most rapid drop south and east of southern James Bay (Schueler and Cook 1980). Pseudacris t. maculata ranges east to Moosonee on southern James Bay. South of the Great Lakes it is replaced by an eastern subspecies, the Western Chorus Frog, P. ¢. triseriata. These forms do not make contact in Ontario (Cook 1964b; Weller and Palermo 1976; and unpublished observations) but do southwest of Lake Superior (Conant 1975). This study presents evidence of introgression in Bufo a. americanus from the central form (B. a. hemiophrys) east to James Bay in western Ontario. This area corresponds to the eastern range of Pseudacris t. maculata. The influence of T. s. parietalis into T. s. sirtalis populations is not known to be as extensive, apparently reaching its eastern limit in extreme western Ontario. These distributions of species and variations may in- dicate past isolation of central and eastern herpetofaunas, followed by the re-establishment of contact and subsequent eastward dispersal of some forms and gene flow from others primarily from the central region into the eastern region. Most of the regions discussed here were covered by glacial ice during at least four successive con- tinental glaciations (Flint 1971) each of which would have obliterated northern faunas and floras over their present ranges. It is generally assumed that much of the present northern fauna and flora was displaced southward in response to changing climatic conditions south of glaciation and sur- vived south of the ice sheet in ‘‘gross refugia’’ (Lindroth 1969). The most recent continental glaciation, the Wisconsin, reached its maximum extent about 18,000 years ago (Prest ef al. 1969). Bryson and Wendland (1967), Flint (1971) and Terasmae (1973) present maps reconstructing climate and vegetation during the Wisconsin max- imum which depict the entire glacial margin bordered to the south by tundra and boreal forest. However, recent evidence indicates the possibili- ty of a region in Nebraska during peak glaciation of drifting sand dunes which may have been treeless or even vegetationless (Wright 1971). If the area between these dunes and the glacier margin was occupied only by tundra, the Boreal Forest would be broken into eastern and central segments, assuming the southern edge of the dune areas was occupied by prairie vegetation. (Mengel (1970) invokes this possible break in discussing zoogeography of bird taxa.) Present central Boreal forms show a gradation of southern limits from Rana sylvatica which does not penetrate grassland habitats (Cook 1965b) to Pseudacris triseriata maculata which is well adapted to northern grassland but is replaced by related species in the southern prairies (Conant 1975). All would be ex- cluded from a severely arid barrier. If the Boreal Forest was broken in central North America, cen- tral Boreal forms would be isolated along the eastern side of the Rocky Mountains and adjacent areas. Rana sylvatica and B. (a.) hemiophrys now have disjunct populations in southeastern Wyo- ming and the present ranges of Rana pipiens, Pseudacris and T. s. parietalis also include this area (Conant 1975, Stebbins 1966). 70 Although not a primary barrier, post-glacial Lake Agassiz which covered much of southern Manitoba (Mayer-Oaks 1967) may have had an ef- fect on the present distribution patterns. It has been postulated by Léve (1959) that the vegeta- tion which first covered the lake bed after Lake Agassiz drained was grassland and another grassland interval occurred in the Hypsithermal. During either of these transitional periods, B. (a.) hemiophrys, assuming it had reached the area, may have been able to invade farther east than at present. At least some of what has been inter- preted here as eastward introgression of B. a. hemiophrys genes could be the result of initial B. a. hemiophrys invasion of the area, and subse- quent swamping by B. a. americanus which spread into the area with the invading species from the eastern forest. However, it is unlikely that the postulated post-Agassiz invasion of B. a. hemiophrys was extensive enough to explain the far eastern extent (James Bay) of indications of introgression from B. a. hemiophrys. The east-central faunal split may have occurred much earlier than the Wisconsin or more than once through successive Pleistocene glaciations. Tihen (1962b) tentatively identified material from the Kansan glaciation in Kansas as similar to B. a. hemiophrys. Tihen (1972) has suggested that B. hibbardi from the Miocene-Pliocene boundary and the Middle Pliocene, and B. rexroadensis of the Upper Pliocene may represent temporal stages in a single population line leading to B. woodhousei. If this interpretation is correct it in- dicates the comparative antiquity of the B. americanus group and the long period available to differentiate distinctive forms. In a survey by Gelbach (1965), the oldest fossil assigned to other members of the B. americanus group is the ten- tative B. a. hemiophrys above. Many cases of interbreeding between mor- phologically distinctive taxa have been explained in terms of comparatively recent human changes to the environment, particularly through land clearing for agriculture (Mayr 1963). In the pre- sent case, the interbreeding sampled is along east- west highways through a relatively well-drained area (Warkentin 1967). Here roadside ditches pro- vide new breeding sites across the transition zone. If call serves to attract females and/or other males to breeding sites (Bogert 1960), then lack of ponds before human alterations at the eastern forest- Aspen Parkland transition could have lessened the chance of a male coming to a pond in the ‘‘wrong’’ vegetation zone and of a female responding to the ‘‘wrong’’ chorus. East-west highways through the region are a comparatively recent development, and even east-west railways date back only about 100 years. However, the evidence for extensive eastern introgression argues against contact being recent enough to have been man-induced. A test of this hypothesis would be in the relatively unaltered area north of Lake Winnipeg where the nature of contact is yet unknown. B. a. hemio- phrys and B. a. americanus are present in the area, and their ranges are in close proximity. Harper (1963) noted what he recognized as calls of B. a. hemiophrys at Wabowden, and Vere Scott (per- sonal communications based on his own observa- tions and those of Leigh M. Nelson and Gordon Shaw) identified B. a. americanus from Warren’s Landing and William River at the north end of Lake Winnipeg, based largely on the diagnostic cranial crests. Field surveys conducted by F.W. Schueler, R.M. Rankin, F.D. Ross, and S.F. Baird in May and June 1980 at the north end of Lake Winnipeg have confirmed the existence there of a hybrid zone similar to that in southeastern Manitoba. Detailed analysis will be published later. Endler (1977) has‘developed a substantial alter- native argument to explain what have often, in the past and in this paper, been regarded as points of post-glacial secondary contact between previously geographically separated populations. He shows that, in theory, such sudden morphological changes could be induced by strongly opposing selective pressures greatly reducing gene flow over an abrupt ecological change. The formation of these sharp breaks in variation are postulated to have been possible even in the short period since glaciation without any previous absolute geographic isolation. In retaining a more tradi- tional (zoogeographic) viewpoint here, I have been swayed by the degree of morphological similarity within B. a. americanus and B. a. hemiophrys populations over their respective ranges, despite the other ecological changes they encounter, as contrasted with the sharp change apparent over the Aspen Parkland-eastern Boreal ecotone in southwestern Manitoba. The more gradual colour and pattern variation exhibited by B. a. ameri- canus populations primarily on a north-south axis, discussed earlier, seems, however, to be more in line with Endler’s models, and requires no postula- tion of previous geographic isolation. 71 7. Taxonomic Significance of the B. a. americanus-hemiophrys Relationship with Respect to the Other Nominal Species of the B. americanus Group The B. americanus group of Blair (1972) consists of B. woodhousei, regarded as an early split from the remaining species, and the largely or com- pletely allopatric B. terrestris, B. americanus, B. houstonensis, B. hemiophrys and B. microscaphus (Conant 1975; Stebbins 1966). B. woodhousei is known to hybridize naturally with all of these taxa except B. hemiophrys. A study of the region where contact between B. woodhousei and B. hemiophrys is probable through South Dakota, North Dakota and Montana is long overdue. B. terrestris and B. americanus may hybridize along a contact zone but the evidence is disputed (Netting and Goin 1946; Neil 1949). B. houstonen- sis in Texas and B. microscaphus in the southwest are disjunct from contact with other taxa in the B. americanus subgroup. B. houstonensis shows some features of both B. terrestris and B. americanus (Sanders 1953) but is commonly assumed to be a postglacial isolate of B. ameri- canus (Blair 1972). Blair (1972 and earlier papers) suggested an east-west split of a B. hemiophrys- microscaphus stock from the eastern B. americanus stock. Sanders (1961) postulated hybridization and introgression between americanus and an invasion of a broad-skulled toad, possibly of the B. valliceps group, as an ex- planation for the divergent osteological features of B. terrestris. A.P. Blair (1955) postulated in- trogression from a B. boreas stock to account for some morphological features of B. microscaphus, although this view was not considered by W.F. Blair (1957b). This study shows that hybridization between B. a. hemiophrys and B. boreas results in a toad with well-separated, low, somewhat B. a. americanus-like crests. A previously defined subspecies of B. woodhousei, B. w. veletus in Texas and Louisiana, is now regarded as the prod- uct of hybridization between B. w. woodhousei, B. w. fowleri, and B. americanus (Conant 1975). W.F. Blair (1972 and earlier papers) has shown that toads of the B. americanus group have a marked ability to hybridize in all combinations in the laboratory, and many F, individuals have proved fertile in backcrosses (Blair 1963). Only B. woodhousei is sympatric to any extent with other species in the complex, and, although hybridiza- tion is common (as in B. americanus and B. w. fowleri), partial temporal isolation and ecological isolation (breeding site and non-breeding habitat) have developed and distinctness is often main- tained over large areas of geographic sympatry. B. woodhousei seems to warrant its status as a species distinct ’ from: the ‘others of “the: B. americanus group by the criteria of Short (1969). In addition, Jones (1973) has presented evidence that reproductive isolation may have intensified in 30 years in one area of sympatry of B. woodhousei and B. americanus. However, Loftus- Hills (1975) has questioned aspects of this analysis. The status of the other species is more proble- matical. The case for conspecific status of B. a. americanus and B. a. hemiophrys, two of the most morphologically distinctive taxa in the group, has been presented. The B. americanus-terrestris rela- tionship has never been sufficiently analysed. B. houstonensis has been regarded as the most weakly distinctive of the group (Blair 1972). The geographic isolation of B. houstonensis and B. microscaphus is not sufficient in itself to warrant separate species status for each. Laboratory hybridization studies (Blair 1972), ecology and the occurence of natural hybridization of both B. houstonensis and B. microscaphus with B. woodhousei, suggest that if either were in contact with other americanus subgroup toads they would interbreed. It may be that the most pragmatic eventual classification to indicate relationships within the B. americanus subgroup will be to regard terrestris, americanus, houstonensis, hemiophrys and microscaphus as either allospecies or megasubspecies of B. americanus (or rather, of B. terrestris which is the oldest name, see Schmidt 1953). A choice between the two would be difficult but because of their similar interactions with B. woodhousei the latter may be best. Either would be consistent with the close relationships noted by many authors from many lines of evidence (Blair 1972) and place them into better evolutionary and zoogeographic perspective with other taxa of North American Bufo. In part, a taxonomic decision on taxa pairs as closely related but distinctive as the pair analysed here is often coloured by the speculative opinion of the taxonomist concerned as to the past events influencing the present situation and eventual out- come of their interactions. Some assume that isolating mechanisms will inevitably develop and/or intensify (Remington 1968) and therefore the allocation of species-status is actually a predic- tion. Others regard any interbreeding as eventually 72 breaking down the distinctiveness of present mor- phological differences with the eventual merging of the two forms. However, some zones of inter- breeding along ecotones are known or assumed to remain for long periods without changing the taxa (Short 1970). The advantage of the Amadon and Short (1976) treatment is that it distinguishes this intermediate distinctiveness of the taxa and avoids the obscurity of conventional subspecies status (Rising 1970). Because of the size and geographic extent of the samples used, the presumed objectivity of the discriminant analysis, and the repeatability of the measurements and scoring systems used, re- sampling in the future should be possible with results directly comparable to those presented. The weights generated should allow precise evaluation in 10, 25, SO or more years of the then current status and perhaps provide an insight into exactly how, and if, the interbreeding and introgression affects future populations of these toads. Obviously, the possible importance of such interactions in evolution could be in either adding additional variation for selection to act upon or as an intermediate step in eventual attainment of complete reproductive isolation. Only long-term studies, probably exceeding one or more human lifetimes will give sound assessment of such situations. Until now, many laboratory techniques have been applied to problems of interspecific classification in toads (Blair 1972). Too little at- tention has been paid to intraspecific variation (for a notable exception see Guttman 1975). Although natural introgression has often been invoked to explain variable results within taxa, insufficient at- tention has been paid to documenting its existence and extent. As Blair (1972) has suggested for inter- specific studies, a multidisciplinary approach to intraspecific relationships involving extensive analysis of geographic variation in cytogenetic and biochemical characters as well as additional mor- phological comparisons of the type presented here are needed. Subsequent to this analysis, Green (1981) re- sampled a portion of the Trans-Canada transect and examined the variation through isozyme electrophoresis. He compared B. a. americanus hybridization with B. a. hemiophrys there with its hybridization with B. w. fowleri in southwestern Ontario. Although clearly obtaining results similar to the morphological analysis presented here, distinguishing that the two pairs are hybridizing to a different extent, and recognizing that hybrid zones are an indication of genetic relationships, Green asserts that each of these taxa should all be treated taxonomically as species. The utility of the subspecies category is lightly treated as is the use of megasubspecies. In arguing that hybrid zones have little, or at best, only indirect influence on evolutionary change, and apparently accepting the principle that recognizable differences between populations demand a species-level distinction be made, he prefers to define a species on the sub- jective grounds of what a researcher can recognize as different. This approach obviously has appeal, particularly as in the present case when it attempts to rationalize the maintenance of taxonomic levels fic) which are familar through long use in the literature. It seems however, too subjective, and an unsatisfactory manner of reflecting the dynam- ics of the relationships. Green has deposited the specimens from his study at the National Museum of Natural Sciences, where they will be measured and ana- lysed by the methods used here, and a direct com- parison of the two approaches will be reported by the two of us when completed. We are agreed that the importance of these studies lies in examining such interactions in depth from many approaches and the question of the taxonomic arrangement one prefers should not obscure this. References Amadon, Dean. 1966. The superspecies concept. Systematic Zoology 15(3): 246-249. . 1968. Further remarks on the superspecies concept. Systematic Zoology 17(3): 345-346. Amadon, Dean and Lester L. Short. 1976. Treatment of subspecies approaching species status. Systematic Zoology 25(2): 161-167. Anderson, Harry E. 1966. Regression, discriminant analysis and a standard notation for basic statistics. Chapter 5, pp. 153-173 in Handbook of multivariate experimental psychology. Edited by Raymond B. Cattell. Rand McNally and Company, Chicago. xxxi + 959 pp. Ashton, Ray E., Sheldon I. Guttman and Peter Buckley. 1973. Notes on the distribution, coloration and breeding of the Hudson Bay toad, Bufo americanus copei (Yarrow and Hen- shaw). Journal of Herpetology 7(1): 17-20. Backus, Richard H. 1954. Notes on the frogs and toads of Labrador. Copeia 1954(3): 226-227. Bigelow, R.S. 1965. Hybrid zones and reproductive isolation. Evolution 19(4): 449-458. Blair, Albert P. 1943. Geographical variation of ventral mark- ings in toads. American Midland Naturalist 29(3): 615-620. . 1955. Distribution, variation, and hybridization in a relict toad (Bufo microscaphus) in southwestern Utah. American Museum Novitates (1722): 1-38. Blair, W. Frank. 1957a. Mating call and relationships of Bufo hemiophrys Cope. The Texas Journal of Science 9(1): 99-108. . 1957b. Structure of the call and relationships of Bufo microscaphus Cope. Copeia 1957(3): 208-212. . 1963. Intragroup genetic compatibility in the Bufo americanus species group of toads. The Texas Journal of Science 15(1): 15-34. Blair, W. Frank, Editor. 1972. Evolution in the genus Bu/fo. University of Texas Press, Austin, Texas. viii + 459 pp. Bleakney, Sherman. 1952. The amphibians and reptiles of Nova Scotia. Canadian Field-Naturalist 66(5): 125-129. . 1958. A zoogeographical study of the amphibians and reptiles of eastern Canada. National Museum of Canada Bulletin 155: 1-119. . 1974. [Abstract] Geographic variation in amphibians and reptiles of eastern Canada. Program of the 54th An- nual Meeting of the American Society of Ichthyologists and Herpetologists. National Museum of Natural Sciences, Ot- tawa. Herpetological Abstracts, p. 2. Bogert, Charles M. 1960. The influence of sound on the behavior of amphibians and reptiles. Jn Animal sounds and communication. Edited by W.E. Lanyon and W.N. Tavolga. American Institute of Biological Science Publication No. 7, Washington, D.C. Breckenridge, Walter J. 1944. Reptiles and amphibians of Minnesota. Minnesota Museum of Natural History, The University of Minnesota Press, Minneapolis, Minn. xiii + 202 pp. Breckenridge, Walter J. and John R. Tester. 1961. Growth, local movements and hibernation of the Manitoba toad, Bufo hemiophrys. Ecology 42(4): 637-646. Brown, Lauren E. and Michael A. Ewert. 1971. A natural hybrid between the toads Bufo hemiophrys and Bufo 74 cognatus in Minnesota. Journal of Herpetology 5(1-2): 78-82. Brown, W.L., Jr., and E.O. Wilson. 1956. Character displace- ment. Systematic Zoology 5: 49-64. Bryson, Reid A. and Wayne M. Wendland. 1967. Tentative climatic patterns for some late glacial and post-glacial episodes in central North America. Pp. 271-298 in Life, land and water. Edited by William J. Mayer-Oakes. Occasional Papers, Department of Anthropology, University of Manitoba. No. 1, xvi +414 pp. Clarke, Arthur H. 1973. The freshwater molluscs of the Canadian interior basin. Malacologia 13(1-2): 1-509. Conant, Roger. 1975. A field guide to reptiles and amphibians of eastern and central North America. Second Edition. Houghton Mifflin Company, Boston. xvii + 429 pp. Cook, Francis R. 1964a. A northern range extension for Bufo americanus with notes on B. americanus and Rana sylvatica. Canadian Field-Naturalist 78(1): 65-66. . 1964b. Additional records and a correction of the type locality for the Boreal Chorus Frog in northwestern On- tario. Canadian Field-Naturalist 78(3): 186-192. . 1964c. The rusty colour phase of the Canadian toad, Bufo hemiophrys. Canadian Field-Naturalist 78(4): 263-267. . 1965a. Collecting and preserving amphibians and reptiles. Chapter V, pp. 128-151 in Methods of collecting and preserving vertebrate animals by R.M. Anderson. Na- tional Museum of Canada Bulletin 69. . 1965b. Additions to the known range of some amphibians and reptiles in Saskatchewan. Canadian Field- Naturalist 79(2): 112-120. . 1967. An analysis of the herpetofauna of Prince Edward Island.National Museum of Canada Bulletin 212: vii + 60 pp. . 1968. Reptiles and amphibians. Chapter 8, Part III, pp. 436-442. in Science, history and Hudson Bay, Volume I. Edited by C.S. Beals. Department of Energy, Mines and Resources, Ottawa. 502 pp. . 1974. [Abstract] Zoogeography of amphibians and reptiles of western Canada. Program of the 54th Annual Meeting of the American Society of Ichthyologists and Herpetologists. National Museum of Natural Sciences, Ot- tawa. Herpetological Abstracts p. 4. . 1977. Records of the Boreal Toad from the Yukon and northern British Columbia. Canadian Field-Naturalist 91 (2): 185-186. . 1978. An analysis of toads of the Bufo americanus group in a contact zone in central northern North America. Ph. D. thesis, University of Manitoba, Winnipeg, Manitoba, 232.pp. Davis, R.G. 1971. Computer programing in quantitative biology. Academic Press, London and New York. Endler, John A. 1977. Geographic variation, speciation, and clines. Princeton University Press, Princeton, New Jersey. ix + 246 pp. FitzGerald, G.J. and J.R. Bider. 1974. Evidence of a relation- ship between age and activity in the toad Bufo americanus. Canadian Field-Naturalist 88(4): 449-501. Flint, Richard Foster. 1971. Glacial and Quaternary geology. John Wiley and Sons, Inc. N.Y. xii + 892 pp. Garman, H. 1892. A synopsis of the reptiles and amphibians of Illinois. Bulletin of the Illinois State University of Natural History (Champaign, Illinois). Volume III, Article XIII, pp. 215-385 + XV plates. Gaige, Helen T. 1932. The status of Bufo copei. Copeia 1932 (3):134. Gorham, Stanley W. 1974. Checklist of world amphibians up to January 1, 1970. The New Brunswick Museum, St. John, N.B. 172 pp. Gelbach, Frederick R. 1965. Amphibians and reptiles from the Pliocene and Pleistocene of North America: A chronological summary and selected bibliography. Texas Journal of Science 17(2): 56-70. Grant, Ronald. 1941. Salientia of northern Pontiac County, Quebec. Copeia 1941(3): 151-153. Green, David M. 1981. Theoretical analysis of hybrid zones derived from an examination of two dissimilar zones of hybridization in toads (genus Bufo). Ph.D. thesis, Univer- sity of Guelph, Guelph, Ontario. 239 pp. Guttman, Sheldon I. 1969. Blood protein variation in the Bufo americanus species group of toads. Copeia 1969(2): 243-249. . 1975. Genetic variation in the genus Bufo: II. Isozymes in northern allopatric populations of the American Toad, Bufo americanus. Pp. 679-697 in: Isozymes, genetics and evolution, Academic Press, Inc. Harper, Francis. 1956. Amphibians and reptiles of the Ungava Peninsula. Proceedings of the Biological Society of Washington 9: 93-104. . 1963. Amphibians and reptiles of Keewatin and northern Manitoba. Proceedings of the Biological Society of Washington 76: 159-168. Henrich, Thorston W. 1968. Morphological evidence of secon- dary intergradation between Bufo hemiophrys Cope and Bufo americanus Holbrook in eastern South Dakota. Herpetologica 24(1): 1-13. Jameson, David L., James P. Mackey and Margaret Ander- son. 1973. Weather, climate and the external morphology of Pacific Tree Frogs. Evolution 27(2): 285-302. Jones, J. Michael. 1973. Effects of thirty years hybridization on the toads Bufo americanus and Bufo woodhousei fowleri at Bloomington, Indiana. Evolution 27(3): 435-448. Lee. Julian C. 1982. Accuracy and precision in anuran morphometrics: artifacts of preservation. Systematic Zoology 31(3): 266-281. Licht, Lawrence E. 1976. Sexual selection in toads (Bufo americanus). Canadian Journal of Zoology 54(8): 1277-1284. Lindroth, Carl H. 1969. The biological importance of Pleistocene refugia. Chapter 1: pp. 7-17 in The Kodiak Island refugium: its geology, flora, fauna and history. Edited by Thor N.V. Karlstrom and George E. Ball. The Boreal In- stitute, University of Alberta, Edmonton. (Ryerson Press, Toronto) xiii + 262 pp. Loftus-Hills, Jasper J. 1975. The evidence for reproductive character displacement between the toads Bufo americanus and B. woodhousei fowleri. Evolution 29(2): 368-369. Logier, E.B.S. 1928. Amphibians and reptiles of the Lake Nipigon region. Transactions of the Royal Canadian In- stitute. 16(2): 279-291. . 1952. The frogs, toads and salamanders of eastern Canada Clarke, Irwin and Company Limited, Toronto xii + 127 pp. Logier, E.B.S. and G.C. Toner. 1955. Check-list of the am- phibians and reptiles of Canada and Alaska. Royal Ontario ie Museum of Zoology and Paleontology. Contribution 41: i-vi, 1-88. . 1961. Checklist of the amphibians and reptiles of Canada and Alaska. A revision of Contribution No. 41. Royal Ontario Museum, Life Sciences Division. Contribu- tion 53: 1-92. Love, Doris. 1959. The post-glacial development of the flora of Manitoba - a discussion. Canadian Journal of Botany 57: 547-585. Martin, Robert F. 1973. Osteology of North American Bufo: The americanus, cognatus, and boreas species groups. Herpetologica 29(4): 375-387. Mayer-Oakes, William J., Editor. 1967. Life, Land and Water: Proceedings of the 1966 Conference on environmental studies of the Glacial Lake Agassiz Region. Occasional Papers, Department of Anthropology, University of Manitoba (1): xvi + 414 pp. Mayr, Ernst. 1963. Animal species and evolution. The Belknap Press of Harvard University Press. Cambridge, Massachusetts. xiv + 797 pp. . 1969. Principles of systematic zoology. McGraw- Hill Book Company, New York. xi + 428 pp. . 1970. Populations, species and evolution: An abridgement of Animal Species and Evolution. The Belknap Press of Harvard University, Cambridge, Massachusetts. xv + 453 pp. McAllister, Don E., Robert Murphy and John Morrison. 1978. The compleat minicomputer cataloging and research system for a museum. Curator 21(1): 63-91. Mecham, William R. 1962. Factors affecting secondary intergradation between two allopatric populations in the Bufo woodhousei complex. American Midland Naturalist 67(2): 282-304. Mengel, Robert M. 1970. The North American Central Plains as an isolating agent in bird speciation. Pp. 279-340 in Pleistocene and recent environments of the Central Great Plains. Edited by Wakefield Dort, Jr., and J. Knox Jones, Jr. Department of Geology, University of Kansas, Special Publication 3 (University Press of Kansas). 433 pp. Neill, Wilfred T. 1949. Hybrid toads in Georgia. Herpetologica 5(2): 30-32. Netting, M. Graham, and Coleman J. Goin. 1946. The cor- rect names for some toads from the eastern United States. Copeia 1946(2): 107. Nevo, Eviatar. 1973. Adaptive color polymorphism in cricket frogs. Evolution 27(3): 353-367. Packard, Gary C. 1972. Inconsistency in application of the biological species concept to disjunct populations of anurans in the southwestern Wyoming and northcentral Colorado. Journal of Herpetology 5(3/4): 191-193. Porter, Kenneth R. 1968. Evolutionary status of a relict popula- tion of Bufo hemiophrys Cope. Evolution 22(5): 583-594. Prest, V.K. and the Surveys and Mapping Branch Geological Survey of Canada. 1969. Retreat of Wisconsin and recent ice in North America: speculative ice-marginal positions dur- ing recession of last ice-sheet complex. Scale 1: 5,000,000. Geological Survey of Canada Map 1257A. Remington, Charles L. 1968. Suture-zones of hybrid inter action between recently joined biotas. Evolutionary Biology 2 (8): 321-428. Rising, James D. 1970. Morphological variation and evolu- tion in some North American orioles. Systematic Zoology 19(4): 315-351. Rowe, J.S. 1959. Forest regions of Canada. Department of Northern Affairs and National Resources, Canada, Forestry Branch Bulletin 123: 71 pp. Sanders, Ottys. 1953. A new species of toad with a discussion of morphology of the bufonid skull. Herpetclogica 9(1): 25-47. Sanders, Ottys. 1961. Indications for the hybrid origin of Bufo terrestris Bonnaterre. Herpetologica 17(3): 145-156. Savage, Jay M. 1960. Evolution of a peninsular herpetofauna. Systematic Zoology 9(3): 184-212. Schmidt, Karl P. 1938. A geographic variation gradient in frogs. Field Museum of Natural History, Zoological Series 20: 377-382. . 1953. A check list of North American amphibians and reptiles. Sixth Edition. American Society of Ichthyologists and Herpetologists, University of Chicago Press, Chicago. 280 pp. Schueler, Frederick W. 1973. Frogs of the Ontario Coast of Hudson Bay and James Bay. The Canadian Field-Naturalist 87: 409-418. Schueler, Frederick W. and Francis R. Cook. 1980. Geographic variation of the middorsal stripe dimorphism in the wood frog, Rana sylvatica, in eastern North America. Canadian Journal of Zoology 58(9): 1643-1651. Schueler, Frederick W. and James D. Rising. 1976. Phenetic evidence of natural hybridization. Systematic Zoology 25(3): 283-289. Shay, Creighton T. 1967. Vegetation history of the southern Lake Agassiz basin during the past 12,000 years. Pp. 231-252 in Life, land and water. Edited by William J. Mayer-Oakes. Occasional Papers, Department of Anthropology, Univer- sity of Manitoba. No. 1. xvi + 414 pp. Short, Lester L. 1969a. Taxonomic aspects of avian hybridiza- tion. Auk 86: 84-105. . 1969b. ‘‘Suture-zones’’, secondary contacts, and hybridization. Systematic Zoology 18(4): 458-460. . 1970. A reply to Uzzell and Ashmole. Systematic Zoology 19(2): 199-202. Sibley, Charles G. 1961. Hybridization and _ isolating mechanisms. Pp. 69-88 in Vertebrate speciation. Edited by W. Frank Blair. University of Texas Press. Smith, Philip W. 1961. The amphibians and reptiles of Illinois. Illinois Natural History Survey Bulletin 28(1): 1-298. Smith, Hobart M. 1974. Conceptual requirements of the subspecies. The Biologist 56(3): 105-110. Stebbins, Robert C. 1966. A field guide to western reptiles and amphibians: field marks of all species in western North America. Houghton Mifflin Company, Boston. xiv + 299 pp. Stewart, K.W. and F.R. Cook. 1970. [Abstract] Distribution of tree frogs of the Hyla versicolor complex in Manitoba. Pp. 27-28 in American Society of Ichthyologists and Herpetologists Fiftieth Annual Meeting. Abstracts. Tulane University, La. 76 Terasmae, J. 1973. Notes on late wisconsin and early holocene history of vegetation in Canada. Arctic and Alpine Research 5(3:' Part: 1): 201-222. Tester, John R., A. Parker and Donald B. Siniff. 1965. Ex- perimental studies on habitat preference and thermoregula- tion of Bufo americanus, B. hemiophrys and B. cognatus. Journal of the Minnesota Academy of Science 33(1): 27-32. Tihen, Joseph A. 1962a. Osteological observations on New World Bufo. American Midland Naturalist 67(1): 157-183. . 1962b. A review of New World fossil bufonids. American Midland Naturalist 68(1): 1-50. . 1972. The fossil record. Pp. 8-13. in: Evolution in the genus Bufo. Edited by W. Frank Blair. University of Texas Press, Austin, Texas. viii + 459 pp. Timken, Richard L. and Donald G. Dunlap. 1955. Ecological distribution of two species of Bufo in southeastern South Dakota. Proceedings of the South Dakota Academy of Science 44: 113-117. Trapido, Harold and Robert T. Clausen. 1938. Amphibians and reptiles of eastern Quebec. Copeia 1938(3): 117-125. Underhill, James C. 1961. Interspecific variation in the Dakota toad, Bufo hemiophrys, from northeastern South Dakota. Herpetologica 17(4): 220-227. Vladykov, V.D. 1941. Preliminary list of Amphibia from the Laurentides Park in Province of Quebec. Canadian Field- Naturalist 55(6): 83-84. Warkentin, John. 1967. Human history of the glacial Lake Agassiz region in the 19th Century. Pp. 325-337 in Life, land and water. Edited by William J. Mayer-Oakes. Occasional Papers, Department of Anthropology, University of Manitoba No. 1. xvi + 414 pp. Weller, Wayne F. and R. Victor Palermo. 1976. A northern range extension for the Western Chorus Frog, Pseudacris triseriata triseriata (Wied), in Ontario. Canadian Field- Naturalist 90: 163-166. Woodruff, David S. 1973. Natural hybridization and hybrid zones. Systematic Zoology, 22(3): 213-218. Wright, Albert Hazen and Anna Allen Wright. 1949. Hand- book of frogs and toads of the United States and Canada. Comstock Publishing Company, Ithaca, N.Y. xiii + 640 pp. Wright, H.E., Jr. 1971. Late Quaternary vegetational history of North America. Chapter 16, pp. 425-464, in The late Cenozoic glacial ages. Edited by Karl K. Turekian. Yale University Press, New Haven. xii + 606. Yarrow, H.C., and H.W. Henshaw. 1878. Report upon the reptiles and batrachians collected during the years of 1875, 1876, and 1877, in California, Arizona, and Nevada. An- nual Report of the Geological Survey of the Territory of the United States west of the 100th Meridian . . . by George M. Wheeler, appendix L: 206-226. Zweifel, Richard G. 1968. Effects of temperature, body size and hybridization on mating call of toads, Bufo a. americanus and Bufo woodhousei fowleri. Copeia 1968(2): 269-285. 2 rE < bs Oo a) ooo ocooooo = N 0 9€ fe) Oo suoutdads jo 1oquUINN Or09 8£09 L009 cLOP cSOP 090P 890P 6P0P €709 7709 1709 LIIL OEITL STIL SSLol eceol sceol cELOI LVS8 SOS8 cOS8 L7S8 I7S8 spr Joquinu onsojeyes OWN C961 C961 7961 6S6l 6S6l 6S61 6S61 6S61 7961 7961 7961 £961 £961 £961 L961 9961 9961 9961 $961 S961 $961 $961 $961 8961 uOIIda[OD AR 87 KRW 87 API €1 ounf ¢€Z7 oune Ge oun, 77 ounf 77 ounf [7 ABW 07 ABW 07 ABW 07 ABW O€ AP OF ABN 67 sulids KRW SZ AeW 07 ABW SZ AeW SO ABW 07 APW 81 KRW LZ AeW 97 oune vO ayep pueyisyins jo ¢ Aemysty uo g [Wy Z'ET] soptw 7g pueyioyins jo ¢ Aemysiy uo q [WY p¢Z] Soll Cp] (jyousn}) premoy jo py Aemysipy{ uo q [WY ¢'OT] Sopru ¢°9 Woq[V sullg WIQ[V IUlIg ¢ Aemysiy uo JoATYy [YS Jo q [Wy 70] soyiw [9 ‘puod Woq[V 2ULIg jo € AemysiH uo A [WY [°L] Softw pp “JOATY []eYS syoolg jo | Aemysty UO MN [Wy g’O] soTtu ¢°0 (jjouin}) syooig jo | AemysiH UO MN [WY €°T] Soytu g°0 (jyouin}) syooig Jo | AemYysIH UO MN [WY CE] Sotw 7°7Z dIOWYICIIS s1OWMyIVS JO | AemysIH uo q [WY L'p] Sop 6°7Z a1oWMYIVNS Jo | AemystH uo q [WY 7'¢] Sopu O°Z Aresy[ea ‘dALIG MAN pue AN ‘1S Yip Jo uonounl Jo q [Wy 9°Z] soptur g°] pue NY [wy Zc] sop 7'¢ ned IS JO MS [WY E'S] Sopru E"¢E uleld AuoIg JO 9] ABmyYsIH uO q [WY gO] SoTIW ¢'0 uleld AuOoIg Jo 9] ABmysIH uO q [WY g°O] Softw ¢’0 (Jjouin}) qeakiiad Jo 7 ABMYSIH UO MS [WY L'6] Sop O°9 yOPseM JO py ABMYSIH UO N [WY 6'¢€] Sop p°7Z TOATY BMY All[eo0'T PoINsBVIW SUOT}OITJOO JO ISI] 7] XIGNAddV b7.901 “STo90I £10601 IV SOI £S0SO0I SSoIIT UGorlk SOIT OToITT LEVITT /OTOLIT icsocll /IVoVIT 1800S 8000S TE VeCcs: iT1o€S ET o€S /0€ 00S L0o01S SOo1IS LS ofS EYES SS VTEoVS 4 if I oVS iS Co8S S€ Sé SI UBMIYOIVYSeS V8 VL v9 vs Vv VE v7 VI e1oqiVv sduydoiuay ‘vp ofng :a[duies sdudI3Joy (A\) 9pniisu0T] pue (N) epniney JOquINN AVeoO7J qa UOMNIIJOD 103110d “WM 0 O€ p66 6S61 APN 97 suIqqiH JO €Z AemysiH uO ¢ [WY p'9] Sop O'p 9606 J AToLb NWI BJOSOUUTIAT SNUDIIAOUD *D ofng :ojdwies VUSIOJOY 0 6 SEE LS6I API LZ ayeT Sjiaeq jo q [Wy 99] sot Ip P0086 ,Z008P nal B10oyeq YON € O€ «dan AaTTWA JOATY stweVsey “Io[sog JO L8Z 0 9 0198 S961 Pun 97 AeMyYstH UO AS [Wy ¢€°Z] Sop py “JoaTY orwreIeT OboSOI .EEoIP OMI SUIWIOA 97 ve 809 0961 Aine 17-07 dAIASIY ISAIOJ SpOOM sdNIdg ‘UOTIDeS IsvOYIION 101066 .Sto6P WS SI 6 7S09 7961 OuNL p yorog 119d 161086 ,ITo0S Wr rl ¢ COS 0961 euNL pT (jyouin}) z € €9Sp 0961 euNL ET uolNqO UO | ABMYSIH JO N [Wy g’O] soptu ¢°O 67066 £000 WE (JjouIn}) seuOo]UI;y 0 02 9S7ZI OL6I 9uNnL 6 JO OT Aemysty uo ¥ [Wy 6°71] Sop O'g TS 0001 .S007S WZ Sed 0 rl 7S7ZI OL6I ung L ayi Jo OT AemystH UO ¢§ [WY [°9T-L°6] Satu OT-9 SToIOL s€P0€S WI eqoiureyyy (jgouin}) € pL 96r8 S961 APIA ST Binyewig Jo | AemyYystH UO J [Wy L*¢] soplu ¢°Z WPTEOT 6700S S6 0 01 S97S 1961 AR 61 0 El LOTS 1961 API 61 ouleld 19d JO N [Wy ¢*pz] sop zo] pue 0 I €97S 1961 APIA 81 AM [Wy 6°Z] sattur gy ‘AaTTeA JoaTY apfaddy no WPESOT ,LEOS S8 9 LE LOOP 6S6I euNnL LO ayeT punod oyejjng JO puo MN SEoSOT ,€P00S SL (jjouin}) L 9€ p09 7961 APIA OT sulyduioy jo | Aemysiy uo MA [WY Ec] sot EE TS 0801 ,p000S S9 (jjouin}) 4 7 ITIL €961 APIA 87 jodieg jo | Aemysiy uo q [Wy p'ET] soplm Eg 90801 ,6S$06P Ss UOIxIOX JO § [WY 0'€Z] I a6 €£0P 6S6I eune OT So Ep] “6 ABMYSIY Ye oye] YORoT] 87ZoCOT ,S000S St e) O Joquinu ajep AWeI0T (M\) epnisu0 7 Joquinyy sudutdods onsojejeo UOTIDIT[OD pue (N) epniney AVVO T JO JOQGUINN OWN PoINsSedU SUOTIDITJOO JO ISI] :(panunuod) | XIGNAddV 78 0 07 886E 6S61 APW OI Z 67 p96E 6S61 dy 97 yieouasoy jo q [Wy OT] sayiw 9°09 9€009 E109 dl puels] preMPA 9OUlId Z 7 67h7Z ES6I AP OL OT[AsaTyINOY G0eL9 i lloSF OS I ps pele! IL6I eung ¢€ SIaIsoOy-sap-de Tloh9 TS o8P Or € ¢ 8S7Z ZS6I oune | 9 0 IS7Z ZS61 API 67 I € 7H7Z ZS6I APIA SZ 0 I Ip7Z ZS6I APN ¥Z I 0 Or7Z ZS61 API €7 Z I 6£77Z ZS6I API IZ I 8ET7Z ZS61 APN 07 ¢ 14 LET ZS61 AP 61 [sa]] 1das] ,.spueys] usaag,, T0990" aCle0S O€ 0 € O87Z TS6I OUNL 7 0 I 6L7Z ZS6I PUNE IZ Z I 8L7Z ZS6I OunL 6] 0 I LL@Z ZS6I une LI I ZI 6977 ZS6l oune TI [sa[] 1dag] 0 tl S977 ZS6I ouns OT <896 L961 Tidy ¢ AjuNOD xassy ‘Ye [BUOTIeN I]9q WIOd Gece alc ty O8 Ayunog iwoy ‘xed € LY 9169 £961 [dy LI [eIOUIAOIg neapuoy jo N [Wy go] sayiw ¢°o ASalS. iGlect OL ZI 68 8669 €961 [dv 0¢ ssulidg peqsjieD jo AM [Wy 9°{] sayiw oT (OCeSE aCCesF O09 Aeg YON 9 rol p90L €961 API LI Jo LJ Aemysty uo gq [Wy 6'9]] saytw CO] SToOL s6169P Os p 9¢ 9L0L €961 APN 61 TWIeZOUDY Axe] 71008 ,9008P Or IUIPXO JO LI 0 rs 8E7ZI OL6I ABW IE ABMYSIH UO AQ [WY 1°6] Saf Q°9 ‘Y90ID JoAvag 85006 670697 O€ 0 6€ 16Lr1 ZL6I ouNE Z doUOSOO| 0 IZ €8Lrl ZL61 oun | JO AN [WY €°6T] Sop OZ] “Yee1D dowry A accOR, -ECols O7@ I 81 L6LY1 ZL61 eune ¢ | daUOSOO|] toler selols Ol O1leJuO fe) Oo Joquinu o}ep AyyeooT (M) 2pnisuoT JoquinN suswiioads onsojeleo UOT}DITJOD pue (N) opniney Ayleoo'T JO JOQuUINNy OWN PoINseIW SUOTISIIJ[OO JO JsIT :(panuyuod) | XIGNAddV 79 Jopuexo[Vv 0 ZI 86011 8961 ouns ¢ ‘Iq JO TT Aemysty jo M [Wy 7'¢€] soytu 0°7Z 17096 ,LEOS II JOpuexo[V @ I€ 66011 8961 aun ¢ ‘I Jo T] Aemysty uo AM [WY 16] Sayiu O°9 97096 =, LE0OS Ol jpouin} Yovog 0 SI LEOIT 8961 APIN SZ puein jo Z] AemystH uO N [WY p'9] Softw Op 17£096 LE OS 6 yIed [etourlAOI dq yorsg puelL ‘soley yied 0 4 L611 8961 auNL [Z q wor [wy 9°[] alu O°] ‘uotsua}xa ucOseT A696 EEs0S 8 yled 0 (47 96ITT 8961 eUNL IZ [BIOUIAOIg YORag puBID ‘Yydeaq IsaM ‘UOOSeT 17V096 =, ££00S L jjouin} yovog 0 O€ peor 8961 APN £7 puein jo 7] Aemysty uo ¢ [WY p9] SaTIW O*p HOSA Sa GGUS 9 AInqJa}UuedS 0 8 ZEOII 8961 APIN bZ Jo 6¢ Aemysiy UO AN [Wy [’8] saytur Q°¢ AIng19}uUedss JO 6¢ ABMYSIH UO AN [Wy [7g] sop Q°¢ Gouin) ¢ €Z ZLOLI 8961 APIN TE yoeog eioiied 1e) 6¢ AeMYsIY, SuOCTe sayoiqd 17£096 ,STo0S ¢ 0 I OOTTI 8961 oun ¢ 0 Il 6£0IT 8961 APIN SZ 0 Z SEOII 8961 APIN SZ 0 I €EOII 8961 APIN bZ JO 7] Aemysip{ uO KN [WY Z'¢] Sopiu Q°Z 0€096 ,9100S 7 AInqgiaqUuedS 0 €] LLOII 8961 API ET JO 6¢ ABMYSIH{ UO MS [WY 7] SalI O°Z 160096) “J 1Ca0S € (jgoutn}) 0 SI ILOII 8961 APIA OF neqry Jo 6¢ Aemysipy{ uO MS [WY 9°[] Saytu O'] b7096 .vIo0S (6 0 IZ €97ZI OL6I auNnf ST UID Jo 6 AeMYsSIH UO ¢ [WY 7] SolwW —'Z 0 I Z9ZTI OL6I eunE CI TWID Jo 6 AeMYSIH UO ¢ [WY Z’¢E] Saw C'Z 0 I 19ZZI OL6I une CI TWID Jo 6 ABMYSIH UO ¢ [WY g'p] sow C'¢ 96.96 ,9£€.06 I JIaSUDL[. ULBYIAON IY] suonels uoNdaT[oD of/ng eqoiuryy Uld}seq 0 OV 67L8 p961 APN 87 IoqieH Woqoponbsnyy 1600€9 LVobD NZ@ 0 Ss 9IL8 p961 ABW LI STASIOM JO | [Wy 9°T] sopra Ov] CToh9 ,S00SP NI BI]ODS PAON I €¢ ITL8 6S61 APN ZI ‘6 ‘8 wo}UOW JO N [Wy 16] saqiu 0'9 LPoh9 ,TTo9P dl YoIMsunIg MON 6 O Jaquinu ayep AVeoo0'T (M) dpnisuo0 7] Joquinyy suoumtoads ansoyeleo UOTIDIT[OD pue (N) spniney AVed0'J JO JOQqUINN OWN POINSBdUI SUOTIDIT[OO JO ISI] :(panunuods) | XIGNAddV 80 II ¢¢ 80011 8961 APIN FI a1uusy jo pp Aemysipyy uo J [WY ¢E*[][] saftw O’Z VTS6 1S 06 16 Se pe II01I 8961 APN FI a1uusy jo pp Aemysipy{ uo q [WY pC] soytw O'p 870S6 1S 06P LC yInowsiYAM JO q ‘TT pue p sAemysIH{ jo 0 p7 E8P8 S961 AP EI uonount Jo p Aemysty uo gq [Wy [97] saytw O'OT WVeS6 aoSobr 97 6b ZION 8961 APIN FI euyq Jo [] AemysIH UO NY [WY gp] soyiu Og 1$$0S6 .SSo6P 4 p 6b 8ZOII 8961 ARI 7Z UPIAIA JO CT ABMYSIH uO J [WY 6°71] Soplw O's 91096 .€S06P v7 0 L LZ011 8961 ABI 7Z UPIAIA JO Sy] ABMYSIH UO J [WY E*[[] Satu O'L 81096 .€S06P v7 UPIAIA JO CT ¢ 67 6ZOII 8961 APIN ZZ AemysiyH uo J [WY gp] Sop Q'E “JoATY uoyYoIg £7096 ,£S06P €7 0 <3 OPOll 8961 API 97 0 € O€OII 8961 API €7 UBIAIA JO CT ABMYSIH UO q [WY 9']] AIW O'] 97096. £S06P ce JOASUDAT UDIALA AY] youuog Np oe] 0 €1 06011 8961 eune fp JO [] AemyYsIy uo ¢ [Wy L*zT] sop [| V0096 ,S000S IZ Inofasneag 8 8P 61011 8961 ABW SI JO pp ABmysiy uo J [WY ¢°7Z] saplw Op] 171096 = .PO00S 07 Inofesnesg 0 L OZOII 8961 APN SI JO pp Aemysty uo q [Wy ¢*pT] saplw 0'6 61096 ,70.0S 61 Inofasneag 0 9] IZOII 8961 APN SI JO pp Aemysty uo q [Wy ¢E'[]] soplw Ovz 10T096 ~— .PO00S 8] Inofasneag Z OL SIOII 8961 APIN SI JO pp Aemysty uo q [WY [°g-p'9] Saytu ¢-p $7096 V0.0 LI Z IZ LIOII 8961 APIN SI Inofasnesg jo N [Wy 9°T] aw OT 1€096 .S000S 9] L ZI OI7ZI 8961 eunfs €7 0 I 60771 8961 AUNL €Z yled [eloulAolg [IIH S.Ppilg 000L6 .LSo6P SI JIOSUDA J, dnofasnvag ay JOUUOG 0 Z £6011 896] euNnL p np oe] jo [|] AeMYsIH UO NY [Wy [°g] soytw o°¢ P0096 6100 vl sles 0 € P6011 8961. ounf ¢ yeoIH Jo yy] Aemysty uo ¢ [Wy 9°[] soft O°] 110096 9700 El sylej 0 SI $6011 8961 eunL ¢ ould Jo [J Aemysipy uo AM [WY 9°]] sop O°] V1096 .PE00S aA 6 re) Joquinu o1ep Aylyeso J (AM) IpnysuoT JoquinN suoutdads ansojejeo UOTDITJOD pue (N) apniney AyyeooT JO JOQUINN OWN poeINnseoW SUOTJS9I[OO JO jSI’T >(ponunuod) | XIGNAddV 81 _ NOMNA TMA MNOAMNMAHN MON TRwABANMNAWA HANA t+ st — ONAN OS SO -9Oo ONO .90,.N SC SO COS =O SC.cnic. Oo oS oS oo © QS) SSS] SS OMANN YO © fo) re) suowitoods jo Jaquinyy CCCCl T7771 61CTI Orstl SI8tl CI8It OI8IT 68LI1 LESCOL Liccl Svc! vOccl 81771 vICcl €IC7l SOCTI 100CI . Tesil 88LI1 BELLU SOLIT 9ELII Lvsil cesil IT81T O6LI1 IvOTT L8v8 vOr9 9809 SECTI vECTI Stcel CECTI Teal Joquinu ansoyejeo OWN OL6I API €Z OL6I APIN €Z OL6I APIN €7 6961 AuNE ¢ 6961 APIA 87 6961 APIN LZ 6961 APN LZ 6961 API SI OL6I PUNE ZI OL6I APIN 7 OL6I APIN 7 OL6I ABW 81 OL6I API 7 OL6I APIN IZ OL6I APIN 07 OL6I ABW 8I 6961 AINE I 6961 ouns ¢ 6961 ABW SI 6961 AP L 6961 APIN € 6961 Idy 61 6961 euny ¢ 6961 euns p 6961 API LZ 6961 APIA SI 8961 APIN LZ S961 APIA PI 7961 OUNL g 7961 UNL g OL6I API IE OL6I ABW IE OL6I ABW TE OL6I APN IE OL6I APN OF a1ep UOTIIITJOD II pue | sAemysty jo uonountl Jo | Aemysipy{ UO AX [WY 9°g¢E] SoTIU O'pZ II] pue | Aemysty jo uonounl Jo | AemystH UO A [WY 6'Lp] SOT 61°67 I] pue | shemysipy Jo uonount jo | AemystH UO M [WY Esp] SaTIW O'OE IT pue | skemysiyH jo uonounf Jo | Aemysty UO AA [WY O'OS] SeTtw O'TE II pue | sAemystyY jo uonounl Jo | Aemysipy{ uo MA [WY ¢€°O¢] Softw CZTE (jgouin}) suuy ‘a1g Jo q [WY 84] SOTIW Q'E ‘JJOUIN] BIDIATY 9xVT JO AA INosNnG Joysry JO MA TT pue |] skemysty jo uonounf jo | Aemysty uo AA [WY 9°0S] Sotw PTE I] pue | skemysty jo uonounf jo | Aemysty UO MA [WY p'6p] SOTIW LOE ({ AemysIpY 18 Jjouin}) susoljnqg (JJOUIN}) I[[TAIOAIN JO 6¢ ABMYSIH UO N [WY gB’p] SOT OE AWyeoo7T POINSBSU SUOTIDIT[O JO ISI] :(ponuyuods) | XIGNAddV 197096 = .6V.06t ce VE 0.96 60 060 I€ £7096 ,VPo6P Ot 165096 8&6P 6c Joasudd][ ADMYBIPY DpDoUdD-SUudAT AY] (M) 29pnitsuo 7] Joquinny pue (N) epniney AYeIOT 82 “AYW[ESOT Sty} WO1J sudUtoads pyZ J9YIO BY} UO paseq dB YOIYM SUBIL ONLI pUv a}eLIBATUN dy} WIJ poqIWO Inq (CE) AIU][BSO] SITY] JOJ DOULIIeA PUL URAL 9109S JUBUIWIOSIp Ul papnypout , I] pue | sAemysiy jo uonount 0 t 66601 8961 AP ZI jo | AemystyH suo AM [WY [OT] Seytw OO L096 ,606P SE I] pue | skemysty jo 0 6 19€S 1961 uns TT uonounfl jo [| Aemysiy uO M [WY p'9OT] saytw 70] 80096 ,6P06P LE (II pue [ sAemysty jo uonounl jo | Aemysipy uo A [WY L°LT] sop OTT =) | ABMYSTH 1 JOATY 0 v7 ZO8TI 6961 API SZ peayuayxoig jo AeMYSIH{ UO g [WY 26] SolIw 0°9 60096 ,6706P 9€ 4 0 O77 OL6I APN €7 ‘ €7 ZO7ZI OL6I AP 81 0 Lx PSs 6961 ounL 9] € 91x 6E8II 6961 eune ¢ 0 Px SESII 6961 eune p 0 Px 97811 6961 ounL Z €1 €7* 60811 6961 API LZ I 0 SO8TI 6961 API 97 0 v 10811 6961 APN €7 0 Z OO8TT 6961 APN €7 0 SI 86LI1 6961 APN 77 I] pue | skemysty jo uonountl Il Ol O8LII 6961 APN FI jo | Aemysty uo M [WY p°LZ] STW O'LT Z LI SZOII 8961 APIN IZ 0 LZ EOI 8961 API PI I] pue | Aemysty jo uonounf jo | Aemysty 0 9 €6601 8961 APW II uo M [WY PLZ] Sot O'LT ‘JoaTY peayuoyoig Joyory jo q ‘TI pue |, skemystpy uorounl | 9€ p8rs S961 ABW EI jo | ABMYSIFR UO M [WY g°¢Z] soyIUI O°9OT [I pue |, sAemysty jo uonountl 0 pL 7909 7961 oun L jo | AemystyH Uo M [WY 0'8Z] Sot PLT ,L1096 ,6706P SE c II L771 OL6I APN $7 0 el p@7ZI OL6I APN PZ II pue | sAemysty Jo uonountl 0 el 807ZI OL6I AP 61 Jo | AeMYsIH UO AM [WY 0°67] Satu OST 81096 ,67.6P ve , IT pue | skemysty jo uonounl 0 rl 9Z011 8961 AP 7Z jo | Aemysty uo A [WY PSE] soft 0'7Z 17096 60.69 I] pue | skemysty jo uonounl 0 (43 p9ES 1961 eune TI jo | AemyYysty uO AM [WY p'9E] SaTtu 9°7Z 07096 67069 €€ 6 Oo Joquinu ojyep Ayl[eo0'T (AM) 2pniusu0 T Joquinyy suswiisads ansoyeieo UOTDIT[OD pue (N) opniney Aeo0'7T JO JOQUINN OWN PoINSBIU SUOT}OI]JOO JO ISI] :(ponunuod) | XIGNAddV 83 ¢ LE 67771 OL6I APN O€ 1OSJOL, JO 6 APMYSIH{ UO C [WY p'9] SalI O*p 60096 ,700.6P (Ns Jol][[9}9T 0 81 6b771 OL6I euNL 70 JO CL ABMYSIH UO § [WY 7 €-9'[] Sop Z-] 0 b 8P77ZI OL6I eunL 0 Jatf[a1aT JO CL AeMYSIH UO ¢§ [WY p'9] Softw O'p 0 L Lv@T1 OL6I euNnL 0 Jat][a1aTJ JO ¢L AeMyYsIH UO ¢ [WY 9°¢] sau ¢*¢ 0 al OPTI OL6I eunL FO JatfjajaT JO CL AeMYSIFR UO § [WY g’p] SoTIW O'E 10L6 S006 6r 0 67 ISTIT 8961 ouNL LT dAIISOY IS9IOF [SUV AN 0 LI OPI 8961 ounL LT SPOOM OY} JO aye] ‘Avg oyesjng “Wo Yyog V10S6 01 06P 8p 0 99 SEIT 896] oun 9] dAIOSay 1S9IO4 0 ¢ eel 8961 ouNL OT a]3UV AAN ‘24k asooy ‘puod duinp asnjoy 07oS6 JTL o6t LY 0 9] 880IT 8961 eunf 10 pueyoiey 0 6 18011 8961 API TE JO OIZ ABMYSIR uO | [WY p'9] sop O*p 81096 ,LT6P or pueyoleyy 0 9 6LOII 8961 APIA TE JO OIZ ABMYSIR uO | [WY 7'¢] Sou O°Z] 07096 ,LT060 SP 0 ZI 8S77ZI OL6I eune ZI 0 pL Proll 8961 APIA LZ yorquiays Jo 7] APMYSIR UO ¢ [WY 6°71] SoTIW Og 17.96 ,8706P a4 JIOSUDA T PUudYy DG ay] yled 0 (43 LLV8 S961 APN ZI [BIOUIAOI ][OY4say AA ‘axe T UOsTey Pus MAS Z ril pS09 7961 OUNL SO DAIOSOY ISOIOT [JOYSoUY AM “Ox uooley 61056 ,1€06t eV (Jjouin}) axe] 0 L 66601 8961 APN EI uogey jo | AemYsIH UO M [WY L*6] SoTiw O'9 LT0S6 PEo6t (64 II pue | sAemysipy jo 0 L L6601 8961 APN EI uonounf jo | AemystH uo ¥q [WY 6°71] SaTiw Og 85056 67060 Iv [I] pue | skemysty jo 0 SE OSES 1961 9UNL OT uonount jo | AeMyYsIy uo q [WY p*] SoTtw O°p 67.S6 60060 Or ¢ 9] LOSTI 6961 APIN 97 IT pue | skemysipy Jo ¢ 43 raunal 8961 APIA LZ uonjount jo | Aemysty uo py [WY 7 ¢] 1: 0'7 9 El 00771 OL6I AW LI 0 i O€8II 6961 aun ¢ Z ¢ S6LII 6961 APN 81 0 9 p6LII 6961 ABW LI 6 II L8LII 6961 API FI 0 vl CLLII 6961 ARI 9 I] pue | skemysiyy{ jo ¢ 89 96601 8961 APIA EI uonount Jo | AemysIH{ UO AA [WY B*p] Soplw O'¢E 85056 67060 6£ o) Oo Jaquinu o}yep Ayyeso'T (M) 29pnitsuo TT Joquinyy suduisods onsojejeo UOTIDATJOD pue (N) opnineyq AqeVI07 JO JOQUINN OWN POINSBIUI SUOTID9T[OO JO ISI] :(panunuods) | XIGNAddV 84 v (64 SSoll 8961 APN 67 JJOUIN} O1Q2|PPIW 3e Z] APMYSIH v 97 LSOII 8961 APIA 67 O1g2|PPIW JO ZI AemysiyH uo q [Wy g’o] aT!WI ¢'O ST0S6.. A0c6¥ 19 (jyouin}) € 69 9SOII 8961 APIN 67 anseids jo Z] AemYysIH uo M [WY 9°]] IW OT] 16£056 ,70.6P 09 jyouin} uonouns 0 SE SSOIl 8961 APIN 67 yinog jo Z] Aemysiy UO MN [WY 9°J] ATW OT LVS6 .V006P 6S Auld JO N “68 pue Z] Skemysiy jo 0 SZ O90II 8961 APIN 67 uonounf jo 7] AeMYsIH uo Y [WY gp] sop OE 1$$0S6 ,90.6P 8¢ 0 OP OLOTT 8961 APIN OF AsUIg JO 68 ABMYSIH UO § [WY 9°¢] sap Cg 0 IZ 69011 8961 APIN OF AduIg JO 68 ABMYSIH UO ¢§ [WY Bp] satu O'¢E 65056 ,10.6t LS Aduld JO N ‘68 pue ZI sAemysiY jo uonounl v 07 €Ssoll 8961 APIA 8Z JO 7] ABMYZIH UO MN [WY p'9] Sati O'p 171096 ETL 06P 9¢ Asuld JO N ‘68 pue Z] sAemystpy jo uonounl 0 ral ISOII 8961 APIN 8Z JO 7] AemysIH UO MN [WY ETT] Sap ove £0096 ,7I06P ¢¢ (JjouIn}) OUISIUaTy 0 8Z S90II 8961 APN OF JO 7] ABmYsIH UO MN [WY QT] TI O'T 60096 90.67 vs 0 6 LOOIT 8961 API OF ZI ABmysIY 1e JoATY IY Jo qs [Wy gp] salu O'¢ 51096 GL Ve6P €¢ 71 AUMYSIY 12 6 6£ S9OIT 8961 APIN 6Z Joary yey Jo Z] Aemysipy uo q [Wy 9°{] aIW O'] 91096 71 06P 7] Aemysty I 9 Z90II 8961 APIN 62 ye JOATY WY JO Z] AemyYSIH uo ¥ isn¢ SLTS96 VC1c6P zs ZI ABMYSIP 18 IOATY p cS SrOll 8961 APIA 8Z yey JO 7] Aemysiy UO MN [WY Bp] saTiu OE JOATY 4 0 Lroll 8961 ABI 8Z yey Jo 7] Aemysiy uO MN [WY [°g] say o°¢ 17096 ,Io6P IS } re) Joquinu o1ep AyleooT (M) opniysu0T JaquinN suoutdads onsojejeo UOTIDIT[OD pue (N) opniney AWeo07T JO JOQUINNY OWN POINSVSUI SUOT}JDIT[OO JO ISI] :(penuyuod) | XIGNAddV 85 (Jjouin}) 8P OIIL E961 API 87 1odeig jo | Aemysiy uo q [wy p'¢Eq] sop ¢°g L00601 .6S.6¢ UBMOIYIIEYSeS snjpusoo ofng I E7S8 S961 API 97 YOPISEM JO pp AemYSIH uO N [WY 6'¢] sopru p°7 CSo£TT 6000S BLO Vy sdiydoiuay X svasoqg ofng 09 O£S8 S961 ABI 87 eoseqeyry Jo Z ABMYSIH UO A [WY Z'¢] sop 0°Z Teokl oe spaiog ofng L 16S8 S961 Ounr €Z Ol $8sg S961 OuNL €Z ‘OD SYOOY “U0IYIOIS JO YSg [wry 9°¢] soprur ¢v¢ V1066 61 06€ sesuey 10SNOYPOOM 1aSHOYPOOM ofng 61 ZLO6SI €L6I SUNS 6] OD YJOJION “ulog suo7T ST008 JPEoCP Olle1uG lajmof lasnoypoom ofng iO Joquinu susWIdads onsoje1es o}ep (AX) 9pnysu07 JO Joquiny OWN uOTIDITIOD Aylpeso0'T/satsadg pue (N) 9pniney PoINSBSUT SUOTIDI][OO JO ISI] :(panuynuods) | XIGNAddV 86 ‘(suswIdeds Cp) UONPUIWIIOSIp [[@ UI pasn sudwioads , (Zp)Z- “1-OP8II 6961 eunf ¢ Z (LE)Z- ‘I-SI8II 6961 APIN 87 Z (9€)I-ZI8II 6961 APIN LZ I [I pue | skemysiy (9€)Z- ‘I-OI8II 6961 APN LZ z jo uonounl jo | Aemysty uo M [WY 6*Lp] Sayiw 61°67 (69)I-LS7ZI OL6I ABW ZI I (OS)p- ‘Z- ‘I-LI@ZI OL6I APN 7Z € I] pue | skemysty (OS)I-SI7ZZI OL6I API 7Z I jo uonounf Jo | AemysIH UO A [WY 66h] Sel O'OE (OS)Z- ‘I-817ZI OL6I APN ZZ ye (6r)9- 01 7-PIZZI OL6I ABW IZ ¢ I] pue | skemysty (6€)€- 01 “T-TEST 6961 oung ¢€ € jo uonounf jo | AemYsI_ UO MM [WY 66H] Sow OIE (Op)Z- ‘I-ZE8II 6961 UNL p Z I] pue | skemysty (pZ)Z- ‘I-O6LII 6961 ABW SI Z jo uonounf jo | Aemysipy uO MA [WY ¢'O¢] SaTtwW CZ IE (OL)Z- ‘I-8SZZI OL6I eunE ZI Z yorquiais jo Z] Aemysipy uO ¢ [WY 6°71] SalI C's (19) I-Ip7@ZI OL6I euns | I Ba[UZ[D Jo CL AemysIH UO N [WY 9] Soll Op, (Z9)b- ‘Z- ‘I-€P7ZI OL6I ouns | € Bg[UZ[DH Jo cL AemystH UO N [WY 9°{] aIW OT] (€9)I-pPZZI1 OL6I euny Z I Ba[UZ[D Jo CL AemysIH UO N [WY g’O] Salt CO. (as)s- “g= /I-SEsal OL6I AP IE € (8$)Z-PE7ZI OL6I APN IE I (8¢)8- (LS)€- 91 I-€€7ZI OL6I API IE (LS)Z- ‘I-ZEZZI OL6I API IE Z (9S)ZI- 01 9- “(SS)S- 01 I-1EZZI OL6I APIN O€ al JJOUIN} ST[IAIOAIN JO 6¢ APMYSIPR UO N [WY gp] Sop Q'E (pS)L- 91 1-67ZZI OL6I APIN OF pp 10jSJOL JO 6§ ABMYSIH UO ¢ [WY 9] SaTIW Ops (L9)8I- ‘9I- ‘SI- ‘(99)PI- 1 OI- *($9)6- 01 S- “(p9)p- 01 I[-6h7ZI OL6I euns p LI Jatjaiay JO ¢L AemyYysiy uo ¢ [Wy 7°¢-9']] Soy 7-T, (p9)€- 01 I-9PZZI OL6I UNL pF € Joq[a19T JO ¢L Aemysiy uo ¢ [Wy gp] SopIW O'Ex (0L)8- 91 I-€97ZI OL6I euns ¢] 8 TWID Jo 6 AemystH UO g [WY Z'¢€] SoTIW OZ (69) 1-9S7ZI OL6I PUNE 6 I jgOUIn} seUO}UI Jo OT ABMYSIYR uO | [WY 6°71] Sat O'S (69)L- ‘(89)9 01 I-ZSZZI OL6I eunL ZL L Seq ayL Jo OJ Aemysiy uo ¢§ [WY ['9I-L'6] SOTIW OT-9x sdaydoiuay ‘pv ofng siaquinu ansojeies DIN 21eq suoutoads Al[eo0 TJ | JO JOqUuINN ‘so]duies aduatJajal [eoIsSO;OYdIOUW! 9Yy] WOIJ popnpoxs eqopUPIA UIO}SeIYINOS UI SdtITBdO] AUTOS ae Sa;duIYS ddUdIAJaI [Teo dy] UI pepnyoUyT [Teo pue Asojoydious Jo suostieduiod ay} UI pasn ud0q SARY PUP PdINdUIOD d19M ‘IDAIMOY ‘SIONS JUBUIUWILIOSIP sy] “sIsATeue ;eoIZo;oydiour ay) UI PEPNISUI 3q 0} SUdTDAdS MAJ 00} PSUTLJUOD PoUTe}gO 319M SBUIPIOIII YIYM WO] Sotjeso] May Y “¢€ pue Z INST ul poddew pue ‘apnisuoy] pue opnine] yum | xipueddy ul poloquinu o1e solji[eoo] pue siskjeue [eoIsojoydiow dy} UI pasn OsTe d1aM PpapsOdaI sUdUTIOadS IsSOPY ‘Ioquinu ode} Aq d[Ilf pled XOpUI UO SadUdIDG [eINJeN JO WNasN dy} Je UOTIDAS ABO[OJOdIDH 94) Ul OINN 294) 1® aff UO st UsWIDads Yovd IOJ posn eyed ‘IoquINU oNZoje}ed ade} 9y} Aq (sIsayjUeIed UT) paMmoTOJ ‘1O] IVY) UTYIM Joquinu (s)uswtdads JenprIAIpU! dy] pue Yysep eB Aq paMoOT[O} ‘Iaquinu 0] (OJIN) epeued Jo sumMasny [euONeN oy) oJe UdAIZ sJoquINU sNnSoTRIeD ‘POP1OIII IIOM S][Vd ZUIPIIIQ IJBUI YOIYM JO} SUOTIDIT[OO puv SdT]ITeOO] JO SIT TT XIGNAddV 87 (Sp)EZ- 01 SI- “(LP)LI- 9} ZI- ‘(Op)TI- 01 9- “(Sp)S- 91 I-ZO7ZI (Zp)Z- “I-p88II (Ip)p- 91 T-6E8TI (Op) p- 0} I-8E8TI (SE)E- 02 I-9781T (9¢)0I- (SE)6- 01 b “(PEDE- ‘T-608TT (TE)I-1O8TT (O€)Z- ‘T-008TT (pS)I-877Z1 (pS)€- 01 I-LZZZI (€S)9- 01 €- ‘(7S)Z- “I-p77ZI (€S)I-97ZZI (ZS)I-€77ZI (€p)Z- ‘I-7Z61T (Z7$)6- 91 I-ZZZZI (1S)p- 0} I-17ZI CIS)C- “6c Sloquinu onsoje1ed OWN ‘PapIOdEI d1aM sTTed SUIPseIq oTeW YOY OL61 6961 6961 6961 6961 6961 6961 6961 OL6I OL61 OL61 OL6I OL61 6961 OL6I APN €Z OL6I APIA EZ OL61 APIN €Z ABW 81 aun 9| aun ¢ aune > aune Z ABW LZ ABW €7 ABW €7 ABW 7 KRW 7 KRW 7 ABW $7 ARI $7 aunt 67 neq Z9 Ofng deIpsulisjuUl [e210] €Z Z v 4 € 6 I IIT pue | sAemystHy 6 Jo uonounf jo | AemystH UO AM [WY PLZ] Sow O'LT I € I] pue | skemysiy 9 jo uonounf jo [| AemystH Uo A [WY 0°67] STW O'sT I II pue | sAemysty I jo uonounf jo | AemystH UO A [WY 9°O€] STW 0'6] I] pue | shemysty Z jo uonounf jo | Aemysty uo AA [WY BEE] SaTIW O'1Z suoiojndod ajpipauiiajuy suauisads Al[eoo 7] as sduydomuay ‘dD ‘g [e1OL 6 p I] pue 7 Aemysiyy Z jo uonounf jo | Aemysty uO AQ [WY 9°gE] SalI O'PZ jo 19qUNN JOJ SUOT}D][OD pue SaT}ITBdO] JO ISI] :(ponuNnuod) J] XIGNAddV 88 (suswideds Cp) UOBUIWIIOSIp [[e@d UI popnyour , €7Z ofng [e101 OS Ssnuvodidawp “Dp Ofng |e1OL (09)ZI- 01 L- “(6S)9- 91 €-8EZZI OL6I APN TE OI YUpxO Jo LI Aemysty uo M [WY 16] SaTTW 0'9, O1IBJUQ (ph)8- 01 [-007ZI OL6I ABW LI 8 (9Z)S- 01 I-S68II 6961 APIA SI ¢ (SZ)S- 01 I-P6LTI 6961 API LI ¢ (PZ) I-€6LT1 6961 API OT I (PZ) I-ESLTI 6961 API ZI I I] pue | sAemystyH MEDIEI~ “6- “i> “O- “E> “Z- “I-SLEEI 6961 APN 9 L jo uonounf jo [| Aemysiy{ uo M [WY Q'p] Satu O'Ey (pE)S- 1 I-LO8TI 6961 APIN 97 ¢ IT pue [ sAemysip (LZ)E€- 91 I-96LTI - 6961 AP 61 c jo uonounf jo [| Aemysiy{ UO AM [WY g°p-Z'¢E] SoTIu €-7, I] pue | skemysiy jo uonounf jo | AemystH uo MA [Wy L*LT] sept TT =) (€E)S- O} I-ZO8II 6961 API SZ ¢ [ ABMYSIY 1e IOATY peoyusyolg jo q [Wy /’6] soytw Q'9 snuvoidawup ‘p ofng SloquINU aNSoTRIedD DIN a1eq suaurtoads Ayyeoo FJ jo Joquiny ‘PeplOdII 319M S][VO SUIPSIG [BUI YOIYM IOJ SUOTIDATJTOO pue SdTIITBOOT JO ISIT :(panunuod) IT XIGNAddV 89 Sima er ies yore cal Atal” Sn i as me ee ; * {EER- oF I- CORT ae (eM & - { vyewdail ‘® evi his eaaesl tn 7 cs | \ ae eee cers a yo satreat to t reward eh, 5 | a cE ce oe | | A ee a aes poate e ‘ ; (TSH- Or 41-0077 i = Cae t vali vf E 6 oa viene te j vyewdatH no al fend ees buf 8 ii oan £2 oe < mo . : ee m as Mm : . {hg j2- =] “Osi! GAOT. yal ae ? a AE bn { by af : , ities ont ae ee ; i & . 7 - ; ' / = -_ 7. x 7 : " ) “SjEt- @ es & fe 2 j- 4¥) i Che | ek 2) , ; iG iiitt shift ig I vie wrtis: ae ne vi (ena ‘a, ) zglion o.eF"- ee “7 Ate L- x, yes | am + fe ae iy | (eo)i-Eet ii — @aet wie ot | If tne agit ee el a aa B\EL FOOT? hae rhe =f ; $ ; ! 1 yew: a ” od 7 ’ = t «= es é ‘ Yi Ps >. 4 oe 7 : | | (EDR on beerT! Say veh TA ; | iC) hom my i . 4% ; iS Lab ye teers CMie«/ Zi. 2 . sae ae (db )8- Hi gece! OO? veh4 T1 / ¥ fag copa gor — Ta ae ‘ ns : i c es ik) sav. bal” acs b Ar Oe ae a : : o~ a a RR a a Gabe 3 | ibe) to TI veweail so W fad 58) esterg.ee 9 oe A - 4 , 9 . ’ — ; ‘7 Fie A b “ * a 3 hg Ad geo (-wan | | ve wnNDMeMte 6 OY dgioT - = 1 ( eci- Pe ~ 1 ae : [ bo ald . 2 ¢rr , . ay a . wt o : ropa AG oT a a = 4 =" — si an | - —— —— - es a - a on eee ~ — A A A in pa _ _ ee 7 Sy yi 7 ee tae ee t¥) Sraduimenies re Terr 7 “i . 7 may rs a j ie v . 4 : “i x ; Aum . - ee % ei stig s ei ; : vn, | Te he \ ' : tite &y i Gtr) unr i's) at; . tt Pam ) pithy ; ; ) . ; ia The O'R, eoire iy iS . 1) ae f sem atigs 7 n aie Tae, Bs on Ge G Ce eoniy ore i 4, - ~ — i . ~ s é 7 ; chads! atunpuasiagy ia ; a ( hg i « \ - Aas 4 1! egy win a Q tito i* > i wu 7 Be Sew a” wa | 7 ~ . a4 ¥ fuss; i 3 : 7 o (frat oe ote : - ee ms Lata + Sry a9 pray i Oe re eet ee ee | eae - = fa x = “s> “4 , = P > 7 ; ; a Apscms oe ; oe a : rie * 2 ; ' @ = § ~~ = a pay ee a . } | 5% ? 7 : aa Za a 7 “ . a ee ae : | _ poe I ve 7 5 “Panes, “— > *2) ory _ a ' : ' ; . 4 = 5 Say a = . » we = <7 95 2: as oy - : ¥ x e48 tan ; . SS Ee SS REPT BF By APO RASS 9p Yer ae « pa, o's a ae ‘ * i? ae is he = Aad ‘ : Pai ve a Pp a ré - . i oo a : _ ad Pay = é ne ” oy ~ re > ore ses! ; aA ) on ci ‘4 “ee, 7-* sare o, “a as Fina ad in a < - s\ Ae =~ it NV