—- ey

HARVARD UNIVERSITY & Library of the Museum of

Comparative Zoology

hg eel

a is u os

(

ee en Sees ie

i)

fii, i i

Ree Ath,

oF

tas

Nes CN

y

nanan nv

i ea

j

owe any

Mae ie

THE CANADIAN FIELD-NATURALIST

Volume 91 1977

THE OTTAWA FIELD-NATURALISTS’ CLUB

OTTAWA CANADA

, Fae f heet PS ARIE

NOV 291977

The CANADIAN®*"*~ PIPLD-NATURALISTE

Published by THE OTTAWA FIELD-NATURALISTS’ CLUB, Ottawa, Canada

Volume 91, Number 1! January-March 1977

The Ottawa Field-Naturalists’ Club

FOUNDED IN 1879

Patrons Their Excellencies the Governor General and Madame Jules Léger

The objectives of this Club shall be to promote the appreciation, preservation, and conservation of Canada’s natural heritage; to encourage investigation and publish the results of research in all fields of natural history and to diffuse information on these fields as widely as possible; to support and co-operate with organizations engaged in preserving, maintaining, or restoring environments of high quality for living things.

Members of Council*

President: Ewen C.D. Todd Elisabeth Beaubien Jo Ann Murray Vice President: Roger A. Foxall WJ. ri coe Ney

: : : A.W. Duga era yen Recording Becta: A.J. Erskine . Berl mR, Grew Roger Teves Corresponding Secretary: A. J. Erskine Daleaubite Stanley M. Teeple Treasurer: Pamela J. Sims H.N. MacKenzie C.G. van Zyll de Jong

Diane McClymont

* This Council is in office until the Annual Business Meeting in January 1977.

Correspondence: Address to The Ottawa Field-Naturalists’ Club, Box 3264, Postal Station C, Ottawa, Canada K1Y 4J5

The Canadian Field-Naturalist

The Canadian Field-Naturalist is published quarterly by The Ottawa Field-Naturalists’ Club with the assistance of contributions from the National Research Council of Canada and The Canadian National Sportsmen’s Show. Opinions

and ideas expressed in this journal are private and do not necessarily reflect those of The Ottawa Field-Naturalists’ Club or any other agency.

Editor: Lorraine C. Smith Assistant to the Editor: Donald A. Smith Book Review Editor: J. Wilson Eedy

Associate Editors

C. D. Bird W. Earl Godfrey George H. La Roi E. L. Bousfield Charles Jonkel David P. Scott Francis R. Cook J. Anthony Keith Stephen M. Smith A. J. Erskine Charles J. Krebs Robert E. Wrigley Copy Editor: Marilyn D. Dadswell Business Manager: W. J. Cody Production Manager: Pauline A. Smith Box 3264, Postal Station C Chairman, Publications Committee: C. G. van Zyll de Jong Ottawa, Canada K1Y 4J5

Subscriptions and Membership

Subscription rates for individuals are $7.00 per calendar year. Libraries and other institutions may subscribe at the rate of $15.00 per year (volume). The annual membership fee of $7.00 includes club publications. Subscriptions, applications for membership, notices of changes of address, and undeliverable copies should be mailed to: The Ottawa Field-Naturalists’ Club, Box 3264, Postal Station C, Ottawa, Canada K1Y 4J5.

Second Class Mail Registration No. 0527 — Return Postage Guaranteed.

Back Numbers

Most back numbers of this journal and its predecessors, Transactions of The Ottawa Field-Naturalists’ Club, 1879-1886, and The Ottawa Naturalist, 1887-1919, may be purchased from the Business Manager.

All material intended for publication should be addressed to the Editor:

Dr. Lorraine C. Smith, Department of Biology, Carleton University, Ottawa, Ontario, Canada K1S SB6

Cover: Arctic hares (Lepus arcticus monstrabilis) photographed by Gerald R. Parker on 5 August 1973 at Mokka Fiord, Axel Heiberg Island, Northwest Territories. See article on page 8.

The Canadian Field-Naturalist

Volume 91, Number | January-March 1977

Distribution and Abundance of Waterfowl Wintering in Southern Quebec

AUSTIN REED and ANDRE BOURGET Canadian Wildlife Service, Environment Canada, P.O. Box 10 100, Ste-Foy, Quebec G1V 4H5

Reed, Austin and André Bourget. 1977. Canadian Field-Naturalist 91(1): 1-7.

Distribution and abundance of waterfowl wintering in southern Quebec.

Abstract. In January and February of 1974, 1975, and 1976, surveys of wintering waterfowl were conducted through most open-water areas of southern Quebec. These surveys indicated the presence of at least 171 000 ducks, mostly diving and sea ducks in the estuary and Gulf of St. Lawrence. Inland freshwater areas, mainly in the Montreal region, supported many Common Goldeneye ( Bucephala clangula), Common Merganser (Mergus merganser), and Black Duck (Anas rubripes). The most abundant ducks on the estuarine portion were Oldsquaw (Clangula hyemalis), Common and Barrow’s ( Bucephala islandica) Goldeneyes, and Black Duck. In the gulf, Common Eider (Somateria mollissima), Oldsquaw, Common and Barrow’s Goldeneyes were abundant. The area of the estuary and gulf is of international importance as a sea- and diving-duck wintering ground. Further study and close surveillance are required owing to the birds’ apparently great vulnerability to oil pollution and habitat change in a very rigorous climate.

Résumé. Aux mois de janvier et février 1974, 1975 et 1976, des inventaires de sauvagine ont été effectués dans la partie sud de la province de Québec. Au moins 171 000 canards, représentés principalement par des canards de mer et des canards plongeurs ont séjourné durant l’hiver dans l’estuaire et le golfe du Saint-Laurent. Dans les eaux douces de l’intérieur, surtout dans la région de Montréal, on a recensé bon nombre de Garrots communs (Bucephala clangula), de Bec-scies communs (Mergus merganser) et de Canards noirs (Anas rubripes). Dans lestuaire, les espéces observées en plus grand nombre furent le Canard kakawi (Clangula hyemalis), le Garrot commun, le Garrot de Barrow ( Bucephala islandica) et le Canard noir. Dans le golfe, l’ Eider commun (Somateria mollissima), le Canard kakawi, le Garrot commun et le Garrot de Barrow étaient presents en abondance. Nos données prouvent que les régions de l’estuaire et du golfe sont parmi les quartiers d’hiver les plus importants du monde, surtout en ce qui concerne les canards de mer et certains canards plongeurs. I] serait souhaitable que lon poursuive des études et qu’on exerce une surveillance étroite sur les populations hivernantes de sauvagine a cause de leur grande vulnérabilité a la pollution par l’huile et en raison des changements qui surviennent dans un habitat en climat trés rigoureux.

Several species of water birds overwinter in southern Quebec, principally along the St. Lawrence river, estuary, and gulf. Heavy ship traffic during the winter and expanding urban and industrial development along the shores are posing an ever increasing threat to those birds and to their habitats. The paucity of published information prompted us to assemble recent un- published reports and to conduct surveys to document the distribution and abundance of waterfowl inhabiting that area during the winter.

Study Area and Methods

In early February 1974, 1975, and 1976, we attempted to cover all open-water areas in

southern Quebec by ground and/or aerial counts.

Regional differences in accessibility, habitat type, and species composition prevented the use of a standard survey procedure for the entire study area. We subdivided the area into 10 zones based on these regional differences (Figure 1).

The 1974 survey was carried out from 31 January to 20 February, in 1975 from 3 to 6 February, and in 1976 from | to 12 February. Ground crews (P. Blais, A. Bourget, H. Briard, G. Chapdelaine, P. Dupuis, G. Fortin, P. Lamothe, M. Laverdiere, D. Lehoux, P. Rancourt, A. Reed, L.-G. de Repentigny, J. Rosa, J.-P. Savard, and G. Tremblay)

2. THE CANADIAN FIELD-NATURALIST

HED, / Z i; (fe LACHINE RAPIDS <7 QUEBEC é Yi,

LAKE ST-LOUIS

1G0_ MILES

60 KM

F—— ornteans ! ISLAND

ONTARIO

FIGURE |.

Vol. 91

LEGEND Numbers of ducks, all s

SEPT -ILES

ae NEWFOUNDLAND

Cot

fice E ISLANOS

PRINCE

Cuan ISLA

SE 4

Southern Quebec, showing major duck wintering sites (dark circles) and survey zones (outlined and identified by

large letters). The Montreal area is shown in a larger scale on the inset at the upper left of the figure. The survey zones are (A) Ottawa River, (B) Montreal, (C) Eastern Townships, (D) Lake St. Peter, (E) St. Lawrence estuary, (F) North shore, Gulf of St. Lawrence, (G) Gaspésie, (H) Baie des Chaleurs, (1) Matapedia, (J) Lake St. John.

methodically scanned all open-water areas with telescopes and binoculars in all zones except part of zone F. Dupuis and Tremblay conducted aerial surveys that covered mainly zones E and F. In most flights a Cessna 337 was flown at an altitude of approximately 60 m parallel to, and about 70m seaward from, the edge of the shoreline of permanent ice. At some coastal sites flights were conducted up to 15 km from shore to check for offshore flocks of sea ducks.

Supplementary observations were available from previous years for the Montreal area and the north shore of the estuary, mainly since 1964. Files of the Canadian Wildlife Service and the Quebec Wildlife Service yielded unpublished results of various winter surveys conducted from 1952 to 1963, mainly in the Montreal area. The winter season was considered to extend from | January to 28 February.

Results

The combined results of the 1974, 1975, and 1976 surveys are presented in Table | and Figure 1. We believe that we consistently under-

estimated the bird populations because (1) exhaustive surveys were not possible over sucha vast and partly inaccessible area, (2) many of the species inhabited offshore areas and were dif- ficult to detect from aircraft (see Stotts and Olson 1972), and (3) cold air temperatures and ice conditions reduced the efficiency of the ob- servers. For these reasons, the largest number of birds recorded in each zone over the three winters can be considered as the best estimate of that population; we have presented those maxi- mum counts in Table 1. The data, in summarized form, are plotted in Figure | to show major concentration areas. Those results, and previous data, form the basis for the description which follows for each area.

(A) Ottawa River

Most of this area was frozen over in winter; only a few natural pools and areas below a hydroelectric dam remained open. In 1974 (no counts in 1975-76) three species of ducks occurred but only in small groups: Black Ducks (Anas rubripes), Common Goldeneyes (Buce-

1977

REED and BOURGET: WATERFOWL WINTERING IN SOUTHERN QUEBEC

TABLE |—Estimated numbers of aquatic birds that overwinter in the St. Lawrence River system, Quebec. Each number represents the maximum count obtained. Surveys were conducted in most regions in February 1974, 1975, and 1976

Regions i) & o ts) = < 3 2 as z oe 65 2 = Species 2 r= ons s = 2 1 ie 8 @ oS Total 2.2 S a0 5 Se aS ge SS gs Ow = oe Ww Za () fea} (OG) 3 642) Black Duck 56 506 71 1906 4 2543 (Anas ribripes) Common Goldeneye 160 2908 164 8328! 979! 332 111 12 8 13002 (Bucephala clangula) Barrow’s Goldeneye l 4 1394 869 260 19 2547 (Bucephala islandica) Bufflehead l 2 69 l 73 (Bucephala albeola) Oldsquaw 6451 2411 46782 48 55692 (Clangula hyemalis) Surf Scoter 185 185 (Melanitta perspicillata) Common Eider 9 91035 505 91549 (Somateria mollissima) Common Merganser 16 5300 196 19 5 5) 7 2 5550 (Mergus merganser) Red-breasted Merganser 2 91 74 98 23 l 289 (Mergus serrator) Miscellaneous | YP 12 3 l 89 duck species?

Sub total — Ducks 234 8787 451 18270 95554 47983 206 24 10 171519 Great Cormoran 21 91 112 (Phalacrocorax carbo)

Black Guillemot 78 57 42 177 (Cepphus grylle) Total 234 8787 451 18348 95554 48061 339 24 10 171808

'May include some B. islandica.

2Miscellaneous duck species by order of importance are as follows: Ottawa River, Mallard (Anas platvrhynchos), Montreal, Mallard, Pintatl (Anas acura), Canvas- back (A ythya valisineria), Redhead (Avthyva americana), Lesser Scaup (Avthva affinis), Ring-necked Duck (Avhva collaris), American Wigeon (Anas americana). Eastern Townships, Mallard, Hooded Merganser (Lophodvies cucullatus); Estuary, Harlequin Duck ( Histrionicus histrionicus), Mallard: North Shore, Black Scoter

(Melanitta nigra).

phala clangula), and Common Mergansers (Mergus merganser).

(B) Montreal

This region had large areas of open water, principally at the inlets and outlets of its two largest bodies of water, Lake St. Francis and Lake St. Louis. The Lachine rapids were partic- ularly important. The Montreal region posses- sed the most important winter-bird concentra- tions in freshwater areas in Quebec. The

Common Merganser was the most abundant species in the Montreal region; more than 3500 birds spent the winters of 1974 and 1975 in open water below the Cornwall hydroelectric dam, feeding on fish which had passed through the turbines. The Common Goldeneye was second in importance and the Black Duck third.

(C) Eastern Townships All areas south of Lake St. Peter and east of Montreal were included in this zone.

4 THE CANADIAN FIELD-NATURALIST

Open-water areas were found along fast-flowing streams and rivers in this rolling countryside. In contrast to the ice-free areas of the St. Law- rence, which were relatively consistent in size and location from year to year, open water was highly variable in extent and location within and between winters. No large concentrations of ducks were encountered. Common Mergansers, Common Goldeneyes, and Black Ducks occur- red regularly.

(D) Lake St. Peter

This zone had no appreciable areas of permanent open water and no waterfowl were observed.

(E) Estuary

This portion of the St. Lawrence had large expanses of open water throughout the winter. Of particular importance were large sections of tidal flats near the mouth of the Saguenay River where tides and river currents prevented the formation of permanent ice. Along most of the south shore, water areas suitable for aquatic bird use were frozen over.

This vast area included some of the most important wintering sites for waterfowl in the province, especially along the north shore of the St. Lawrence. The most important con- centration of wintering Black Ducks in Quebec was found on the tidal flats near the mouth of the Saguenay River, where more than 75% of the Black Ducks counted in 1974 were found. The Common Goldeneye was the most widely dis- tributed duck on the north shore of the estuary and was the most abundant species. The Old- squaw (Clangula hyemalis) was also important but it was difficult to appraise its distribution and abundance because it often frequented areas too far from shore to be seen by ground crews and because it took flight early and dispersed rapidly at the approach of an aircraft; it was undoubtedly more abundant than our figures indicated. The Barrow’s Goldeneye ( Bucephala islandica) was also encountered frequently. Buffleheads (Bucephala albeola) occurred reg- ularly (although in small numbers) only near the mouth of the Saguenay River. Two other species of aquatic birds also occurred regularly: the Red-breasted Merganser (Mergus serrator) and the Black Guillemot (Cepphus grylle).

Vol. 91

(F) North Shore of the Gulf

This region extends from Pointe des Monts to Blanc Sablon on the Quebec-— Labrador border (about 800 km) and includes Anticosti Island. Because of the vastness of the area and inaccessibility of certain portions, its coverage cannot be considered exhaustive. The inner part from Pointe des Monts to Matamek was surveyed both from the ground and from the air; east of Matamek was surveyed only from an aircraft. Ice conditions are generally heavy in this region although the area of Natashquan and the eastern tip of Anticosti Island is character- ized by less severe conditions (Simpson 1973).

Sea ducks represented the most impor- tant component of the wintering population of aquatic birds. The Common Eider (Soma- teria mollissima) was recorded along the outer north shore of the gulf; most birds of this species were seen in the Jacques-Cartier passage in areas of slushy ice. The Oldsquaw and Goldeneyes (Common and Barrow’s combined) were also abundant.

(G) Gaspésie

Ice conditions in this zone were highly variable. Water areas were completely covered by large sheets of floating ice when inshore winds prevailed. The eastern tip of the penin- sula, however, was more consistently ice-free; most large flocks were observed there.

The Oldsquaw was by far the most abundant bird species observed in 1974 and 1976; most birds of this species were in Baie de Gaspé, which was clear of ice. In 1975, that site was choked with ice and only a small number of Oldsquaws was observed in the entire zone. Similarly, our counts of goldeneye (Common and Barrow’s) varied from year to year, ap- parently reflecting different conditions of ice cover.

(H) Baie des Chaleurs

This large bay was characterized by heavy ice conditions (Simpson 1973), but oc- casionally strong west winds opened up large expanses of the bay. Despite different ice conditions (1974 heavy, 1975 and 1976 light) few birds were observed. Common and Barrow’s Goldeneyes, Oldsquaw, Red-breasted (Mergus serrator) and Common Mergansers, and Great

1977

Cormorants (Phalacrocorax carbo) were en- countered regularly in all years.

(1) Matapedia

Many sections of the river remained open through the winter, but were not heavily used by wintering waterfowl. The Common Goldeneye was the main species encountered.

(J) Lake St. John

The only open-water areas in this zone were found on fast-flowing sections of rivers, principally the Ashuapmuchuan and the Mis- tassini which flow into the lake. Very few birds were present in this area in 1974 (no census in 1975 and 1976). Common Goldeneyes and Common Mergansers were the only species observed.

Discussion

The abundance and diversity of the wintering population is remarkable for an area withsucha harsh and rigorous climate. In winter waterfowl require open water that can offer an abundant food supply and suitable resting areas (Nilsson 1972). Several species can be accommodated in southern Quebec because there is a wide variety of habitats, ranging from inland-freshwater to coastal-saltwater, which are kept open by tides, winds, and currents. Food is available in the form of aquatic organisms, which are notably abundant in intertidal and sub-littoral zones in the St. Lawrence estuary and gulf.

Of the nine species of waterfowl which wintered regularly in this region the Common Eider, the Oldsquaw, and the Common Golden- eye were, in that order, the most numerous. Large flocks of eiders were present in saltwater along the north shore of the gulf, particularly off Natashquan, Baie Johan-Beetz, Havre St- Pierre, and around Anticosti Island. Flocks of several thousand Oldsquaw were found in salt and brackish waters in Baie de Gaspé, the estuary, and along the north shore of the gulf. Common Goldeneyes occurred in fresh, brack- ish, and saltwater habitats, the largest con- centrations being located in the vicinity of the mouth of the Saguenay River (north shore of the estuary) and in the Lachine Rapids near Mont- real.

The great abundance of Common Goldeneye

REED and BOURGET: WATERFOWL WINTERING IN SOUTHERN QUEBEC 5

and Oldsquaw in the estuary and gulf was heretofore unrecognized (Bellrose 1976; Johns- gard 1975). Similarly, Barrow’s Goldeneye was believed to be relatively rare on the east coast of North America (Bellrose 1976; Johnsgard 1975; Palmer 1976; Hasbrouck 1944): our data in- dicate that the St. Lawrence estuary and gulf represent a stronghold of an unexpectedly large population.

Our counts in January and February (1974-1976) indicated an annual wintering pop- ulation of about 171000 ducks, despite in- complete coverage. Clearly our eider estimates must be very low owing to the inaccessibility of many of the known or suspected haunts of this bird; the same applies to the Oldsquaw, with the added complication of its known ability to escape detection by aerial observers (Stotts and Olson 1972). The estimates of goldeneye from coastal areas are undoubtedly low as well. Although precise adjustments cannot be made at this time, it seems likely that the area covered must harbor at least a quarter of a million ducks and it is conceivable that half a million could be involved.

Clearly the most important part of the study area was the estuary and gulf of St. Lawrence (Zones E, F, G, and H), which accounted for 94% of the ducks recorded. Some areas of the gulf which were not covered by this study also support important populations of wintering ducks. Average winter populations (1972-1974) on Prince Edward Island were 3070 Black Ducks, 1540 goldeneye (apparently almost all B. clangula), 1450 mergansers (species not indi- cated), and small numbers of Oldsquaw (Bate- man, M. 1974. Mid-winter aerial waterfowl surveys of the Maritime Provinces. Unpublished report, Canadian Wildlife Service, Sackville, N.B. mimeo. 13 pp.). The Magdalen Islands support a few hundred Oldsquaw (A. Smith, personal communication), while Newfoundland is an important gathering area for Common Eiders (Gillespie and Learning 1974). Unfortu- nately there are too many gaps in the data to allow estimation of the winter duck population of the entire gulf.

Other areas along the Atlantic coast of North America are also important wintering areas. Various data from aerial surveys in Nova Scotia,

6 THE CANADIAN FIELD-NATURALIST

New Brunswick (Bateman, Joc. cit.) and the Atlantic Flyway (winter surveys coordinated by the United States Fish and Wildlife Service) for selected species are presented in Table 2 along with comparative data from the present study. Although those more southerly areas harbor large numbers of surface-feeding ducks it is evident that the St. Lawrence is of great importance to sea and diving ducks.

Comparison with European counts is more difficult because a more complete coverage is obtained there by using a network of ground observers (Atkinson-Willes 1969). In terms of total ducks present, however, few areas of equivalent size harbor as many ducks as the St. Lawrence. For example, a comparison of our figures with those from the Baltic Sea, indicated by Atkinson-Willes (1969, Figure 2, p. 105), suggest wintering duck populations of similar magnitude. Also, much of the North Sea has fewer ducks than does the St. Lawrence (Atkinson-Willes 1969; Milne and Campbell 1973). That area of the North Sea encompassing Denmark, the Netherlands and northern Ger- many, however, has a much larger overwinter- ing duck population. The latter area is probably the only European site of greater importance than the St. Lawrence to diving and sea ducks (Atkinson-Willes 1969; Joensen 1974). Clearly the St. Lawrence must be considered one of the major duck wintering areas of the North Atlantic.

Voleoi

Industrial and urban development along the St. Lawrence pose a constant threat to the birds and to their habitats. Port facilities for super- tankers have been proposed for the estuary and gulf, which would increase the likelihood and gravity of oil spills which could have disastrous effects. In the past two years, 102 cases of oil pollution have been reported in these regions; it is only through good luck that none has had serious effects. In the Lachine Rapids near Montreal, a hydroelectric dam has been pro- posed which, through major changes in the hydrography and biology of the area, could lead to drastic reduction in the Common Goldeneye population. Some habitat changes may lead to an increase in the numbers of some species. The proposed Lachine Rapids dam might increase the numbers of Common Mergansers, as has occurred further upriver near Cornwall. The Lake St. Peter area, which presently has no open water, may eventually serve as a new wintering ground if warm water effluents from nearby nuclear power plants at Gentilly prevent the freezing over of expanses of the St. Lawrence in that area.

This investigation has permitted the identi- fication of the more important areas of the system and provided an approximate estimate of numbers of the various species of wintering birds, but further work is required. In partic- ular, the surveys along the remote outer north shore cannot be considered as exhaustive.

TABLE 2—Summary of counts of selected species of wintering ducks on the Atlantic coast of North America

Nova Scotia and United States}

New Brunswick2

Species St. Lawrence! Black Duck 2500 (Anas rubripes)

Common Goldeneye 13000 (Bucephala clangula) Oldsquaw 55700 (Clangula hyemalis) Common Eider 91500

(Somateria mollissima)

11800 304500 4400 68200 850 12200 2100 72600

'This study, based on maximum of mid-winter counts 1974-1976 (from Table |, rounded to nearest 100).

*Bateman (1974)—Jan. and Feb. 1974 (Atlantic and Fundy coasts).

‘Unpublished results of mid-winter waterfowl surveys in the Atlantic Flyway, coordinated by the United States Fish and Wildlife Service. Average 1964-1973.

1977

Further work is proposed to obtain more accurate counts of sea ducks in the estuary and gulf, to document further the present status of the Barrow’s Goldeneye in winter, and to clarify the racial status of Common Eiders (two races, Somateria mollissima dresseri and S. m. bor- ealis, are. known to overwinter in the gulf (Ouellet 1969, 1975; Gillespie and Learning _1974)). Also it is hoped that a survey can soon be undertaken to cover the entire gulf. The con- tinued and increasing threats from industrial and urban develcpment provide ample justifica- tion for close surveillance and further study.

Acknowledgments

We are extremely grateful to H. Ouellet, National Museum of Natural Sciences, Ottawa, to M. Lepage, Quebec Wildlife Service, Mont- real, and to Y. Lafleur, Parks Canada, for providing data from unpublished reports. Thanks are due to the many wildlife tech- nicians, pilots, and volunteers who participated in the censuses, often under very difficult con- ditions. The Ministry of Transport kindly permitted us to place an observer on board its helicopter during a lighthouse inspection in the gulf in 1974. H. Ouellet, H. Boyd, and J. Bryant made valuable comments on the manuscript.

Literature Cited

Atkinson-Willes, G. L. 1969. The mid-winter distribution of wildfowl in Europe, northern Africa and south-west Asia, 1967 and 1968. Wildfowl 20: 98-111.

REED and BOURGET: WATERFOWL WINTERING IN SOUTHERN QUEBEC 7

Bellrose, F.C. 1976. Ducks, geese and swans of North America. Stackpole Books, Harrisburg, Pennsylvania. 544 pp.

Gillespie, D. I. and W. J. Learning. 1974. Eider numbers and distribution off Newfoundland. /n Canadian Wildlife Service waterfowl studies in eastern Canada 1969-73. Edited by H. Boyd. Canadian Wildlife Service Report Series Number 29. pp. 73-78.

Hasbrouck, E. M. 1944. The status of Barrow’s Goldeneye in the eastern United States. Auk 61: 544-554.

Joensen, A. H. 1974. Waterfowl populations in Denmark, 1965-1973. Danish Review of Game Biology 9(1). 206 pp.

Johnsgard, P. A. 1975. Waterfowl of North America. In- diana University Press, Bloomington, Indiana. 575 pp.

Milne, H. and L. H. Campbell. 1973. Wintering sea-ducks off the east coast of Scotland. Bird Study 20: 153-172.

Nilsson, L. 1972. Habitat selection, food choice, and feeding habits of diving ducks in coastal waters of South Sweden during the non-breeding season. Ornis Scandi- navica 3: 55-78.

Ouellet, H. 1969. Les oiseaux de I’Ile d’Anticosti, province de Québec, Canada. Musée national des sciences naturelles, publications en zoologie Numéro |. 79 pp.

Ouellet, H. 1975. Contribution a l’étude des oiseaux d’hiver au Parc National de Forillon, Québec. Revue Géogra- phique de Montréal 29(4): 289-304.

Palmer, R. S. (Editor). 1976. Handbook of North Ameri- can birds. Volume 3, Waterfowl (Part 2). Yale University Press, New Haven, Connecticut. 560 pp.

Simpson, W. 1973. Gulf of St. Lawrence water uses and related activities. Lands Directorate, Environment Canada, Geographical Paper Number 53. 20 pp.

Stotts, R.S. and D.P. Olson. 1972. An evaluation of waterfowl surveys on the New Hampshire coastline. Journal of Wildlife Management 36(2): 468-477.

Received 12 May 1975 Accepted 21 November 1976

Morphology, Reproduction, Diet, and Behavior of the Arctic Hare (Lepus arcticus monstrabilis) on Axel Heiberg Island, Northwest Territories

GERALD R. PARKER Canadian Wildlife Service, Box 1590, Sackville, New Brunswick EQOA 3C0

Parker, Gerald R. 1977. Morphology, reproduction, diet, and behavior of the arctic hare (Lepus arcticus monstrabilis) on Axel Heiberg Island, Northwest Territories. Canadian Field-Naturalist 91(1): 8-18.

Abstract. Fifty-one arctic hares (Lepus arcticus monstrabilis) were collected at Mokka Fiord, Axel Heiberg Island, Northwest Territories: 35 in summer 1973 and 16 in late winter 1975. Adult female weights averaged 4.5 and 3.9 kg for the summer and winter periods respectively; male weights, 4.1 and 4.0 kg. There were significant decreases in the weights of the heart and kidneys between summer and winter samples, as well as an overall loss in body weight. In 1973 the peak of births was believed to have been approximately 20 June. The average litter size, as determined from counts of corpora lutea was approximately five. Breeding occurred about | May; the period of gestation was approximately 50 days. Arctic hares displayed great diversity in their summer diet, but fed mainly on arctic willow (Salix arctica), Dryas integrifolia, and grasses. Arctic willow was the main species consumed at all seasons, and made up 95% by weight of the winter diet. The abundance of willow, a light covering of winter snow, and broken terrain providing adequate escape cover may explain the high densities of arctic hares on parts of Axel Heiberg and Ellesmere Islands.

The arctic hare (Lepus arcticus) is found in Canada north of the treeline to the northernmost point of land on Ellesmere Island, Northwest Territories, and also on the rock-strewn plateaus and mountains of eastern Newfoundland. In Greenland it is common on most of the ice-free coastal region.

In North America studies of L. arcticus have been limited to short notes on their distribution (Bergerud 1967; J. G. Inder. 1972. Arctic hares on Brunette Island. Typewritten report to New- foundland Wildlife Service. 3 pp.; Watson 1954), taxonomy (Handley 1952; Howell 1936; Nelson 1934), predation (Tener 1954), and natural history (Walkinshaw 1947). Bonnyman (1975) made observations on the behavior and habitat use of hares on the Fosheim Peninsula, Elles- mere Island in the summer of 1975, but detailed information on the biology of the arctic hare is absent from the literature. Arctic hares occur in unusually high densities and often form herds of several hundred or more individuals on parts of Ellesmere and Axel Heiberg Islands. In the summer of 1973 I observed and collected speci- mens of L. arcticus monstrabilis during a study on the feeding habits of muskoxen (Ovibos moschatus) and caribou (Rangifer tarandus pearyi) at Mokka Fiord, Axel Heiberg Island,

and returned to Mokka Fiord in late winter 1975 for further observations and collections.

Study Area

Mokka Fiord is close to the northern limit of arctic hare range (Figure 1). The study area is a valley 6 to 12 km wide; a mountain range is to the west and a lower series of gypsum pierce- ment domes on the east drop abruptly to the waters of Mokka Fiord and Eureka Sound. The valley floor is approximately 200 m above sea- level. The gypsum domes present a region of sparsely vegetated gravel- and rock-strewn ridges and slopes, giving way to an interspersion of gravel ridges and meadows in the valley proper. Meadows are restricted to the glacier-fed streams originating in the mountains to the west.

The slopes west of the valley support Salix arctica, Luzula nivalis, Poa spp., Alopecurus alpinus, Arctagrostis latifolia, Dryas integri-

folia, Saxifraga spp., and a variety of other

forbs. The gravel ridges in the valley support a wide variety of forbs and several grasses; Carex stans, Eriophorum spp., and mosses dominate the meadows. The gravel- and rock-strewn slopes east of the valley support only occasional tufts of grasses (Poa spp., Puccinellia spp.), Carex rupestris, Kobresia myosuroides, and

LOTT

PARKER: ARCTIC HARE ON AXEL HEIBERG ISLAND 9

300 Miles os

Kilometre

| FIGURE 1. The location of Mokka Fiord, Axel Heiberg Island, Northwest Territories.

scattered mats of Salix arctica. Occasional forbs include Papaver radicatum, Oxyria digyna, Saxifraga oppositifolia, and Dryas integrifolia. Lichens are common but not abundant, the dominant species being Cetraria nivalis, C. cucullata, Thamnolia vermicularis, and Parmelia spp.

The mountains of Axel Heiberg Island and eastern Ellesmere Island present a barrier to moist air from the south. This topographical peculiarity contributes to less cloud cover and higher temperatures in summer on western Elles- mere and eastern Axel Heiberg Islands than in most of the Arctic. At Eureka on northwestern Ellesmere Island, mean annual snowfall is 37.5cm and total precipitation only 6.8 cm (Thompson 1967). Mean daily winter tempera- tures are among the coldest in the Canadian Arctic; the average for November to March is —26.9°C.

Methods

Hares were collected from 3 July to 3 August 1973 and from 23 March to 6 May 1975. They were shot at close range (5 to 20 m) witha .22- caliber rifle. Most specimens were shot in the neck to avoid damage to the skull or internal organs. Each specimen was weighed on spring scales and standard body measurements were taken (Anderson 1965). The heart, lungs, full stomach, liver, and kidneys were weighed on a beam balance. Stomach samples were preserved in an alcohol-formalin-acetic acid solution (AFA) for future analysis. Reproductive tracts were also preserved in AFA. The skulls and mandibles were retained, and a small collection of skins was made from the 1975 collection. All material was deposited with the Canadian Wildlife Service, Atlantic Region, Sackville, New Brunswick. General observations on hare behavior were recorded whenever possible.

10 THE CANADIAN FIELD-NATURALIST

In the laboratory, basic cranial and mandib- ular measurements were taken for each speci- men, the specifications conforming with those by Banfield (1961) for caribou. Ovaries were sectioned for a count of corpora lutea. Ovaries were dehydrated and embedded in paraffin, and by use of a rotary microtome, were sectioned in units of 10 uw thick. Every tenth section was mounted on a slide and stained with Weigert’s haematoxylin. External measurements of each ovary were taken in addition to the maximum length, breadth, and depth (number of sections in which structure appears X 10 4) of each corpus luteum.

Differences between measurements for com- ponents of the sample were tested for signifi- cance using a simple /-test.

The diet of hares was determined by the identi- fication of plant epidermal tissue in stomach samples. The preparation of plant reference slides and key and the method of stomach sample analysis is described by Parker et al. (Parker, G. R., B. Campbell, and M. Gauthier. 1976. Estimating the diet of Arctic herbivores by rumen and faecal analysis. Canadian Wildlife Service Report CWSC-2021. Ottawa. 143 pp.).

In contrast to the herbaceous-dominated summer stomach samples, the winter diet con- tained a large amount of unidentifiable woody material. As the only woody plant of any con- sequence is the arctic willow, most of the woody component of the diet was assumed to be that species. Twigs of Dryas integrifolia and the roots of forbs may also have comprised an unknown portion of the non-herbaceous diet.

To quantify the importance of herbaceous and

Vol. 91

woody forage in the winter diet, a wet sample of 15 to 20 g was washed through a series of sieves of the following dimensions: 1.18 mm; 850 yp; 425 yu; and 300 uw. Samples of material from each sieve were examined for presence of herbaceous material. Only the sieve of 300 u contained appreciable amounts of herbaceous material; the others contained woody fragments almost ex- clusively. Contents of the 300- sieve were then analyzed for plant species composition similar to the summer samples. The weights of material in all sieves were totalled and divided into the weight of herbaceous material in the 300-p sieve. The latter weight was estimated by multiplying the total weight in the 300-u sieve by the percentage of herbaceous fragments determined from microscopic examination. This method is believed to give a close approximation of the importance of woody plant material in the winter diet of arctic hares.

Results

General Morphology

Thirty-five arctic hares were collected at Mokka Fiord, Axel Heiberg Island in July and August 1973, and 16 from March to May 1975. The summer collection consisted of 10 adult males, 15 adult females, 6 juvenile males, and 4 juvenile females. The winter collection was composed of eight males and eight females. I did not distinguish between adults and young-of- the-year in the winter collection.

Summer and winter weights and measure- ments of adult hares are shown in Table 1. Inthe summer collection adult females were signifi- cantly heavier (P< 0.01) and of greater girth than adult males. Females collected in the winter

TABLE |—The mean weights and measurements, with 95% confidence limits, of adult arctic hares ( Lepus arcticus) collected at Mokka Fiord, Axel Heiberg Island during the periods 3 to 29 July 1973, and 23 March to 6 May 1975

Sex Sample Weight Length Girth Hindfoot Shoulder height Ear size (kg) (mm) (mm) (mm) (mm) (mm) 1973 Female 15 455 +0.7 675.6 + 404 345.6 + 38.4 156.3 + 11.2 3660 +4 42.2 83.6 + 5.4 Male 10 413 +05 669.0 28.8 338.0 + 29.8 154.6 + 8.4 355:5\ = 28:4) 826 e=e5 22 1975 Female 8 3.99 + 0.4 691.2 + 50.1 SWPis) 2 SA) 153.6 + 6.2 388.7 + 52.2 81.9 + 2.4 Male 8 408 + 0.2 690.6 + 38.8 361.9 + 44.1 150.2 + 9.4 386.4 + 35.2 83.0 + 5.7

1977

weighed less than adult females from the summer collection (P < 0.05).

Weights for immature hares collected in July and August 1973 and a projected growth curve for the first six weeks of life are shown in Figure 2. Although adult females were generally heavier than adult males, the juvenile weights suggest that males put on weight more rapidly than females during the first six to eight weeks of life.

Cranial and mandibular measurements were taken whenever the condition of specimen ma- terial permitted. There was no trend evident in seasonal differences in those measurements.

Seasonal weights of selected internal organs of males and females are shown in Table 2. There were significant decreases in mean weights of the heart between summer and winter samples of both adult males (P< 0.05) and adult females (P< 0.05). Both sexes also showed significant decreases in kidney weight from summer to

2.80

(KILOGRAMS)

-80

WEIGHT

-60

.40

Wt. at birth on June 20 100 grams

28 30 1 4 8 12

20 24 = JUNE ———=]

PARKER: ARCTIC HARE ON AXEL HEIBERG ISLAND 11

winter (males, P< 0.05; females, P< 0.001). The liver also showed a seasonal loss of weight although the loss was significant (P < 0.01) only for females. Mean stomach weights of males were significantly (P< 0.001) heavier in the winter than in the summer. In females the mean stomach weight in winter was greater than that of summer but not significantly so. Winter increases could be attributable to a greater intake of woody material. The mean weights of lungs of both sexes were heavier in summer than in winter, although those differences were not significant (P > 0.05).

Reproduction

The ovaries of 15 adult female arctic hares collected in summer 1973 were sectioned and examined for corpora lutea of pregnancy. All females had ovulated and bred the previous breeding season. The mean number of corpora lutea for the sample was 6.5 (SD = 1.7). The mean

ee e MALES

* FEMALES

20 24 28

16 iia 4 8 JULY |— AUGUST —4

FIGURE 2. Weights of juvenile arctic hares collected at Mokka Fiord in summer 1973, and a projected growth curve from

20 June to 8 August 1973.

12 THE CANADIAN FIELD-NATURALIST

Vol. 91

TABLE 2—Mean weights, in grams, of organs from arctic hares (Lepus arcticus) collected on Axel Heiberg Island, Northwest Territories in summer 1973 and winter 1975 (standard deviation in parentheses)

Number of rF Collection Sex and eee Heart Lungs Liver piles Stomach period age (n) Right Left (full) Summer 1973 Adult males 10 50.5 60.6 93.7 13-8) 13.5 156.6 (6.0) (20.7) (14.0) (1.1) (1.0) (19.2) Adult females 15 50.0 62.4 118.6 16.7 16.5 188.1 (5.3) (12.5) (23.8) (2.0) (2.2) (67.4) Winter 1975 Adult males 7 44.8 67.2 86.3 12.3 12.0 PNS)22 (4.4) (10.7) (33.6) (1.6) (1.3) (29.8) Adult females 5 41.7 67.5 77.0 11.2 11.1 200.5 (8.3) (24.2) (28.8) (1.8) (1.8) (25.6)

*Value of m may vary among organs owing to damage during collection. **Weight of kidney minus fat.

number of corpora lutea for left ovaries was 3.3 (SD = 1.5) and for right ovaries 3.2 (SD = 0.4). The difference was not significant. The mini- mum and maximum number of corpora lutea for both ovaries was 5 and !1 respectively. A corpus luteum of pregnancy from a lactating female collected on 19 July is shown in Figure 3.

The ovaries of seven females collected from 23 March to 4 May 1975 were examined for developing Graafian follicles and corpora lutea. The ovaries of one of four specimens collected on 23 March contained numerous developing follicles although the ovaries from the other three contained but few. Developing and mature Graafian follicles and corpora lutea atretica were numerous in the ovaries of two specimens collected on 4 April (Figure 4). The right ovary of a female collected 4 May contained five recently formed corpora lutea (Figure 5), the left Ovary two, suggesting that ovulation had oc- curred about | May.

The ovaries from the one female collected 4 May 1975 were much larger than ovaries of all other specimens. Corpora lutea in ovaries from the 1973 collection measured approximately 2.5 X 1.5 X 3.0 mm; those from the one speci- men at ovulation in 1975 approximately 2.5X 3.0 5.0mm. Although corpora lutea were easily identified in the ovaries of females collected in the summer of 1973, the structures had disappeared from ovaries collected in March to May 1975.

The first young hare was seen on our arrival at Mokka Fiord on 22 June 1973. The brownish- gray coloration of young hares and their habit of remaining motionless behind stones made their detection difficult.

Food Habits

Thirty-five species or genera of vascular plants were identified in 23 stomach samples of adult arctic hares collected at Mokka Fiord during summer 1973 (Table 3). Of the 35 species or genera of plants recorded, 16 were grasses, 13 were dicots, and 6 were sedges. The mean per- centages of epidermal fragments of those classes in the stomach samples were as follows: dicots 53, grasses 41, and sedges 6. Arctic willow was the most common plant species in summer stomach samples, followed by Dryas integrifolia and the grasses Puccinellia Andersonii, Agropy- ron violaceum, and Arctagrostis latifolia. Those five species accounted for 68.5% of the total plant fragments recorded. Only 11.6% of the recorded fragments could not be identified to genera or species, but all fragments could be classified as being grasses, sedges, or dicots.

Twenty-two species or genera of vascular plants were identified in 16 stomach samples of adult arctic hares collected at Mokka Fiord during late winter 1975 (Table 4). Non-woody material made up only 9.2% (SD = 2.9%) by weight of the winter stomach samples. The remaining 90.8% consisted of woody material

PARKER: ARCTIC HARE ON AXEL HEIBERG ISLAND

13

FIGURE 3. A corpus lut-

eum of pregnancy from a lactating arctic hare collec- ted at Mokka Fiord, Axel Hei- berg Island, 19 July 1973. GLC = granulosa lutein cells; TLC =the- cal lutein cells.

FIGURE 4. Developing

Graafian follicles in the ovary of an arctic hare col- lected at Mokka Fiord on 4 April 1975. AF = atre- tic follicle; GF = growing follicle.

Figure 5. A_ recently

formed corpus luteum of preg- nancy of anarctic hare collected at Mokka Fiord on 4 May 1975.

14 THE CANADIAN FIELD-NATURALIST

TABLE 3—Plant species composition of stomach contents of 23 arctic hares (Lepus arcticus) shot in the summer at Mokka Fiord, Axel Heiberg Island, Northwest Territories. Deter- minations from microscopic identification of epidermal

fragments : Mean

Plant species Frequency rage

density + SE* Dryas integrifolia 100 16.6+ 2.44 Salix arctica 96 AXVS) se DS Puccinellia Andersonii 96 16.0+ 2.36 Agropyron violaceum 91 NOGs2)) 255) Arctagrostis latifolia 61 45+ 2.09 Colpodium Vahlianum 61 11+ 0.33 Alopecurus alpinus 61 19+ 0.58 Oxyria digyna 57 29+ 1.34 Draba spp. 57 hiss, O29 Taraxacum spp. Sy7/ 13+ 0.38 Carex stans 48 2.0+ 0.56 Eriophorum spp. 39 10+ 0.48 Puccinellia angustata 35 Ail se OF Carex spp. 26 Pgsyae > || CVAl Saxifraga oppositifolia 22 10+ 0.50 Pedicularis spp. 22 0.6+ 0.31 Ranunculus pedatifidus 17 0.3+ 0.19 Poa spp. 17 0.1+< 0.10 Saxifraga tricuspidata 13 0.1+< 0.10 Polygonum viviparum 13 0.1+< 0.10 Papaver radicatum 9 0.1+< 0.10 Ranunculus spp. 9 <0.1+ £0.21 Puccinellia vaginata 9 0.2+ 0.21 Deschampsia brevifolia 9 <0.1+< 0.10 Luzula nivalis 9 01+ 0.13 Festuca baffinensis 9 0.1+< 0.10 Poa arctica 9 <0.1+< 0.10 Epilobium latifolium 9 0.5+ 0.40 Hierochloe alpina 4 <0.1+< 0.10 Poa glauca 4 <0.1+< 0.10 Pleuropogon Sabinei 4 <0.1+< 0.10 Luzula arctica 4 <0.1£< 0.10 Carex nardina 4 (ise) O)A8} Puccinellia poacea 4 <0.1+<0.10 Festuca brachyphylla 4 <0.1 =< 0.10 Unidentified dicots 96 U7) 35) MNe8ks Unidentified grasses 96 3.5+ 0.44 Unidentified sedges 52 0.9+ 0.23

*Relative density = number of recognized fragments of a species expressed as a percentage of the total number of fragments of all species.

believed to be almost totally arctic willow. Willow also comprised 15.1% of the non-woody material in the winter stomach contents. Dryas integrifolia, Salix arctica, Puccinellia spp., Agropyron violaceum, and Colpodium Vahlia- num together made up 62.6% of the fragments in the non-woody portion of the winter stomach

Vol. 91

contents. The approximate mean percentages of dicots, grasses, and sedges among the non- woody fragments were 70, 21, and 8 respectively.

Parasites

Cursory examination of the carcasses showed no evidence of internal parasites or pathology. Two adult hares collected in the late winter of 1975 were infested with the flea Hoplopsyllus glacialis glacialis, which has been reported from arctic hares in other regions of northern Canada, but these were the first records for Axel Heiberg Island (G.P. Holland, personal communication 1975). Seventeen specimens of the parasite were found on an adult female hare and two on an adult male. The female had scratched off patches of fur and cut the exposed skin with her claws.

TABLE 4— Plant species composition of non-woody stomach contents of 16 arctic hares shot in the winter at Mokka Fiord, Axel Heiberg Island, Northwest Territories. Determinations from microscopic identification of epidermal fragments

: Mean Plant species Frequency errs density + SE*

Dryas integrifolia 100 37.3 + 3.23 Salix arctica 94 15.1 + 2.10 Puccinellia spp. 81 6.3 + 1.58 Agropyron violaceum 81 2.2 + 0.68 Colpodium Vahlianum 75 1.7+0.35 Draba spp. 63 3.1 + 0.98 Oxyria digyna 63 2.3 + 0.63 Eriophorum spp. 56 1.1 +0.38 Carex spp. 50 1.2+0.53 Arctagrostis latifolia 44 1.3 +£0.58 Alopecurus alpinus 44 0.6 + 0.20 Pedicularis spp. 38 0.4+0.15 Taraxacum spp. 38 0.7 + 0.43 Festuca baffinensis 31 0.6 + 0.28 Carex stans 25 0.6 + 0.33 Polygonum viviparum 19 0.5 + 0.38 Hierochloe alpina 13 0.1 + 0.08 Poa spp. 13 0.2+0.15 Ranunculus spp. 6 <0.1 £0.05 Festuca brachyphylla 6 <0.1 £0.05 Saxifraga tricuspidata 6 < 0.1 + 0.05 Pleuropogon Sabinei 6 <0.1 £0.05 Unidentified dicots 100 10.1 = 0.98 Unidentified grasses 100 8.0 + 0.98 Unidentified sedges 81 4.5+0.80 Unidentified monocots 19 0.8 + 0.43

*Relative density = number of recognized fragments of a species expressed as a percentage of the total number of fragments of all species.

1977

Behavior

Adult hares had constructed numerous “scoops” or shallow dish-like depressions in the gravel ridges and slopes, usually facing south or southwest and behind, or sheltered by, a large rock. Hares often used those protected scoops during periods of unfavorable weather. Whether young were born in those scoops 1s not known but that seems likely. The young rapidly disperse, however, to spend the first two weeks of life hiding behind rocks, showing themselves only to nibble on vegetation, or to nurse. Lactating females were scattered throughout the lowlands, the young normally restricting their distribution to the gravel ridges and slopes.

By the third week of life, young hares had assumed their gray summer pelage and were forming nursery bands numbering up to 20 animals. Although adult females were still in the area, and occasionally within a herd of young, nursing was not observed. Young hares collected on 12 July were feeding mainly on vegetation although hardened chunks of milk remained in the stomach. The herds of young hares con- tinued to grow in size throughout July. By early August most hares at Mokka Fiord had formed into several large herds, containing both young and adults, and were moving erratically throughout the valley.

The behavior of hares to human approach varied. Sudden approach, however, usually brought the hare to its hind legs, standing erect on its toes. That position affords the hare a better view and is probably the reason for its use. If the hare is unable to distinguish the object of disturbance and a rapid approach is continued, it usually flees in “kangaroo fashion.” If pur- sued, the hare quickly drops to a quadrupedal position and moves up slope to cover, often assuming an upright stance once high ground is reached, to look back at the object of danger.

Solitary lactating females showed little fear of humans, often allowing an approach of several metres or less before slowly hopping away. Herds of young hares in late July and August exhibited a more nervous behavior, often fleeing for a kilometre or more on first sight of a human. I believe heavy predation on young hares by wolves (Canis lupus) and arctic fox (Alopex lagopus) caused that extreme flight behavior. Once the hares identified their pursuer as being

PARKER: ARCTIC HARE ON AXEL HEIBERG ISLAND 15

neither wolf nor fox, however, one was able to walk amidst them and be completely ignored. The behavior of the herd in winter was similar to that of late summer. Solitary hares in winter generally displayed more fear than in summer, but again individual behavior varied.

The two major predators of the hare at Mokka Fiord were the wolf and arctic fox. Killing was not observed, but both predators are believed to have relied heavily on young hares in the summer of 1973. Arctic fox were often seen carrying young hares to a nearby den. When young hares approached 2 kg in August, and became difficult for a fox to carry intact, the latter often transported the hare carcass in halves.

The remains of a young hare, approximately 10 days of age, was found at the nest of a Snowy Owl (Nyctea scandiaca) in 1973. The scarcity of owls in the study area, however, diminishes it as an important predator of the arctic hare.

Discussion

Arctic hares at Mokka Fiord, Axel Heiberg Island, are heavier than most southern races of the species but comparable in weight to speci- mens from other high arctic locations (Howell 1936) and Newfoundland (3.4 to 4.9 kg) (North- cott 1974). Waterston and Waterston (Water- son, G. and I. Waterson. 1972. Report on wildlife, vegetation and environmental values on the Queen Elizabeth Islands (Ellesmere and part of Axel Heiberg) June 28-August 15. Type- written report for Canadian Wildlife Service, Ottawa) recorded the weight of an adult lactat- ing female on 15 July 1972 at Fosheim Penin- sula, Ellesmere Island as 9 1b (4.1 kg). The mountain hare of Scotland (Flux 1970) and Norway (Walhovd 1965), however, weighs less than Lepus arcticus monstrabilis.

The tendency for females to be generally larger in body size and weight than males is also in agreement with studies on hares in Scotland (Flux 1970) and Norway (Walhovd 1965). The decrease in weight by both sexes between summer and winter samples reflects a general summer-to-winter weight loss apparently typical of all species of hares (Hewson 1968). The weight loss is believed to result from a decrease in the quality of diet and approach of the breeding season, although*‘the influence of juveniles in the

16 THE CANADIAN FIELD-NATURALIST

winter sample cannot be ignored.

The rate of growth of young arctic hares is much greater than for young mountain hares. Both Hewson (1968) and Flux (1970) report a weight gain of 14 g/day for mountain hares under 2 kg. Arctic hares during their first month of life exhibited an average daily weight gain of 45 to 50 g/day.

Flux (1971) refers to a 10% seasonal loss in kidney weight of European brown hares (Lepus europaeus) in New Zealand. I found kidney weights of arctic hares in winter to be 11-12% lower for males and 33% lower for females than summer weights. Dauphiné (1975) noted a greater loss in the weight of kidneys from female than from male barrenground caribou (Rangifer tarandus groenlandicus) in northern Canada, although the difference was not as great as that reported here for the arctic hare. The loss of kidney weight during the winter period, as well as from the other major organs of arctic hares, is probably, as suggested for red deer (Cervus elaphus) by Batcheler and Clarke (1970), a physiological response to a decrease in forage quality and quantity and a reduction in the basal metabolism.

Arctic hares on Axel Heiberg Island have one litter per year, as do all populations of hares north of the treeline. In Norway the mountain hare may have up to three litters per year, averaging two to four young each litter (Wal- hovd 1965), as does the mountain hare of Scotland (Flux 1970). In Newfoundland the arctic hare may produce two to three litters per year (Northcott 1974). The Alaskan hare is similar to the arctic hare in having only one litter per year (Walkinshaw 1947). A decrease in the number of litters per year appears to be offset by an increase in the number of young per litter.

A corpora lutea count of 6.5 suggests the average litter size for hares on Axel Heiberg Island is approximately 5(R. D. Baker, personal communication), slightly less than earlier re- cords of 6 to 7 for hares from Greenland (Pedersen 1926; Manniche 1910, cited by Howell 1936), and 7 to 8 reported for northern Ellesmere Island (Feilden 1877, cited by Howell 1936). On Brunette Island, Newfoundland, J. Inder (per- sonal communication 1976) reported a mean litter size for nine arctic hares of 3.8, the range being from 3 to S.

Vol. 91

I first saw young arctic hares at Mokka Fiord on 22 June 1973. Waterston and Waterston (ibid.) reported that female hares on the Fosheim Peninsula, Ellesmere Island gave birth to young between 29 June and | July 1972. On Greenland, young hares were first reported on 10 June (Walkinshaw 1947). In the Canadian High Arctic it appears that most births of arctic hares occur in the last 10 days of June.

Examination of ovaries from the 1975 collec- tion suggests ovulation occurred approximately 1 May, giving a gestation period of 53 days. If we assume the birth of hares probably occurred several days prior to our arrival, the gestation period would be approximately 50 days, similar to the period of gestation of 50.3 + 1.3 days for mountain hares in northern Scotland (Flux 1970).

Ovarian sections from females collected in the late winter and spring of 1975 suggest that an increase in follicular development begins near 1 April and continues throughout the month. Copulation induces ovulation in hares and the recently formed corpora lutea in the ovaries of a hare collected on 4 May suggests breeding took place near | May. Although structures resem- bling old corpora lutea of pregnancy were present in the ovaries of several females collected in late winter 1975, determining breeding success of the previous year by acount of such structures was not attempted and is not recommended.

Arctic hares showed a wide selection in their summer feeding habits, although willow, Dryas integrifolia, and several species of grasses made up the bulk of their diet. Whereas the new green growth of grasses and forbs composed the bulk of their summer diet, woody material accounted for over 90% of the winter diet, most of which is believed to be arctic willow. Approximately 15% of the herbaceous component of the winter diet was also willow, suggesting the total proportion of willow in the winter diet may approach 95%.

Sedges were not important in the diet of hares at Mokka Fiord. Sedges are largely confined to the low-lying meadows, areas where hares were seldom seen in summer or winter. The elevated and dry gravel slopes, where hares were most often found at all seasons, supported a sparse but rather diverse flora. Willow, Dryas integri-

folia, and a variety of grasses and forbs were the

dominant vegetation, suggesting that availa-

1977

bility rather than selectivity probably dictates the diet of the arctic hare. I believe hares utilized the slopes more for escape and food availability because of a light snow cover, rather than food selectivity.

Information on the feeding habits of arctic hares in northern Canada is all but absent in the literature. Studies of the mountain hare in Sweden (Lindlof et al. 1974), northeast Lapland (Pullianinen 1972), and Finland (Nyholm 1968) stress the importance of willow in the winter diet. Winter habitat of hare in Scandinavia, however, often includes regions which support shrubs and trees. Bonnyman (1975), after observing feeding hares in late May, mentions the importance of willow in the winter diet of hares on Fosheim Peninsula, Ellesmere Island.

Arctic willow is the dominant plant species over much of the Fosheim Peninsula and on eastern Axel Heiberg Island (Inglis, J. T. and C.J. Jonkel. 1972. Ellesmere Island range studies. Canadian Wildlife Service, Ottawa, MS report. 24 pp.). The abundance of that food source, a light covering of winter snow assuring continuous availability, and broken terrain providing escape cover, may explain the abun- dance of hares in those areas in contrast to their relative scarcity over most of the Canadian Arctic.

My observations suggest that the arctic hare may be an important competitor with muskoxen and caribou during the winter period when all three species feed upon willow. Whether such competition reaches a level where one or more of the mammal species is adversely affected has not been demonstrated. Such competition would occur only where arctic hares reach extremely high densities such as on Axel Heiberg and Elles- mere Islands. My observations indicate that caribou and muskoxen are not as species- restricted in their winter diet as the hares. Caribou feed extensively on upland grasses and sedges while the winter diet of muskoxen may contain a high proportion of lowland sedges.

At Mokka Fiord, one herd of 250-300 hares spent the winter of 1974-75 within an area of approximately 35 km2. Their winter movements could be traced in March by distribution of pellets, remains of broken and uprooted willow, and trampled snow. An area used by hares in early winter becomes virtually unavailable for

PARKER: ARCTIC HARE ON AXEL HEIBERG ISLAND 17

further use by hares, muskoxen, or caribou. The loose snow cover is quickly compacted by the hares, and the extreme temperature and wind abrasion create a firm layer which adheres to the ground. Although perhaps only a few centi- metres thick, it becomes extremely hard. In contrast, adjacent areas not trampled by hares build up a deeper covering of snow; a typical late winter profile would show a thick upper crust over a layer of loose ice granules and air pockets. Where the snow has not been disturbed, a muskox can expose an area of vegetation with several blows of its front feet, tossing the broken crust aside and removing the loose ice granules with its hoof or muzzle. Althougha potential for interspecific competition for food appears to exist, muskoxen at Mokka Fiord displayed such a broad selectivity of habitat for feeding that adverse effects of competition were negligible or absent.

A high density of hares must certainly play an important role in the energy-flow cycle in a relatively closed high-arctic ecosystem such as the Mokka Fiord valley. The importance of muskoxen, caribou, and hares in sucha system 1s poorly understood. A systems approach for further research is highly recommended for such unique areas as the Mokka Fiord valley and Fosheim Peninsula on Ellesmere Island.

Acknowledgments

I am grateful to F. Brazeau and R. K. Ross, Canadian Wildlife Service, for their assistance during the collection of arctic hares; to B. Campbell, National Museums of Canada, Ottawa, and J. Maxwell and D. Morton, Cana- dian Wildlife Service, Sackville, for laboratory preparation and analysis of specimen material; and to G. P. Holland, Department of Agricul- ture, Ottawa, for the identification of ecto- parasites. I also extend my appreciation to J. Inder, Newfoundland Wildlife Service, St. John’s, and D. Flook, Canadian Wildlife Service, Ottawa, for helpful suggestions and criticisms of the completed manuscript.

Literature Cited

Anderson, R. M. 1965. Methods of collecting and preserv- ing vertebrate animals. National Museum of Canada Bulletin Number 69. 199 pp.

Banfield, A. W.F. 1961. A revision of the reindeer and

18 THE CANADIAN FIELD-NATURALIST

caribou, genus Rangifer. National Museum of Canada Bulletin Number 177. 137 pp.

Batcheler, C.L. and C.M.H. Clarke. 1970. Note on kidney weights and kidney fat index. New Zealand Journal of Science 13(4): 663-668.

Bergerud, A. T. 1967. The distribution and abundance of Arctic hares in Newfoundland. Canadian Field-Naturalist 81(4): 242-248.

Bonnyman, S.G. 1975. Behavioural ecology of Lepus arcticus. M.Sc. thesis, Carleton University, Ottawa. 35 pp.

Dauphiné, T.C., Jr. 1975. Kidney weight fluctuations affecting the kidney fat index in caribou. Journal of Wild- life Management 39(2): 379-386.

Feilden, H. W. 1877. On the mammalia of North Green- land and Grinnell Land. Zoology 1(8): 313-321; (9): 353-361. (Not seen by author.)

Flux, J. E.C. 1970. Life history of the Mountain hare (Lepus timidus scoticus) in north-east Scotland. Journal of Zoology (London) 161: 75-123.

Flux, J. E.C. 1971. Validity of the kidney fat index for estimating the condition of hares: a discussion. New Zealand Journal of Science 14(2): 238-244.

Handley, C. O., Jr. 1952. A new hare (Lepus arcticus) from northern Canada. Proceedings of the Biological Society of Washington 65: 199-200.

Hewson, R. 1968. Weights and growth rates in the moun- tain hare Lepus timidus scoticus. Journal of Zoology (London) 154: 249-262.

Howell, A. H. 1936. A revision of the American Arctic hares. Journal of Mammalogy 17(4): 315-337.

Lindlof, B., E. Lindstrom, and A. Pehrson. 1974. On activity, habitat selection and diet of the mountain hare (Lepus timidus L.) in winter. Viltrevy (Stockholm) 9(2): 27-43.

Manniche, A. L. V. 1910. The terrestrial mammals and

Vol. 91

birds of north-east Greenland. Meddelelser om Grdnland 45: 1-200. (Not seen by author.)

Nelson, E. W. 1934. New subspecies of the American Arctic hare. Proceedings of Biological Society of Wash- ington 47: 83-86.

Northcott, T. H. 1974. The land mammals of insular New- foundland. Wildlife Division, Newfoundland Department of Tourism. 90 pp.

Nyholm, E.S. 1968. Ecological observations on the rabbit on Krunni Islands in Ii and in Kuusamo. Suomen Riista 20: 15-31.

Pedersen, A. 1926. Beitrage zur Kentniss der Saugetier und Vogelfauna der ostkuste Grdlands. Meddelelser om Grgnland 68: 151-249. (Not seen by author.)

Pullianinen, E. 1972. Nutrition of the Arctic hare (Lepus timidus) in northeastern Lapland. Annales Zoologici Fennici 9: 17-22.

Tener, J.S. 1954. Three observations of predators attack- ing prey. Number 2, Polar wolf attacking arctic hare. Canadian Field-Naturalist 68(4): 181-182.

Thompson, H.A. 1967. The climate of the Canadian Arctic. In The Canada Year Book, 1967. Dominion Bureau of Statistics, Ottawa. pp. 55-74.

Walhovd, H. 1965. Age determination of the hare (Lepus timidus Linne) with data on age and sex ratios, growth and body weight. Papers of the Norwegian State Game Research Institute, Series 2, Number 22. 57 pp.

Walkinshaw, L. H. 1947. Notes on the Arctic Hare. Jour- nal of Mammalogy 28(4): 353-357.

Watson, A. 1954. Observations on some mammals in Cumberland Peninsula, Baffin Island, in 1953. Canadian Field-Naturalist 68(4): 56-60.

Received 23 April 1976 Accepted 11 August 1976

Life History Observations of Three Species of Snakes in Manitoba

PATRICK T. GREGORY

Department of Zoology, University of Manitoba, Winnipeg, Manitoba R3T 2N2

Present Address: Department of Biology, University of Victoria, Victoria, British Columbia V8W 2Y2

Gregory, Patrick T. 1977. Life history observations of three species of snakes in Manitoba. Canadian Field-Naturalist

91(1): 19-27.

Abstract. Since little is known about the natural history of plains garter snakes, red-bellied snakes, and green snakes, especially in the northern parts of their ranges, various data on these three species were collected over a 34-year period in the Interlake region of Manitoba. As the ecology of the red-sided garter snake in this area is considerably different from elsewhere in its range, similar geographic variation might be expected for the three species studied here. Few differences are in fact obvious; diets, activity cycles, growth, and reproductive features are all basically similar to those in other parts of the species’ ranges. A comparison of the two garter snake species would be of interest for future study because in Manitoba these species apparently exhibit similar growth and reproductive potential, as they do in other parts of their joint range, but have rather

different annual cycles of activity.

Considerable information has been published recently concerning the natural history of the red-sided garter snake (Thamnophis sirtalis parietalis) in Manitoba. These publications have covered activity both at overwintering sites (Aleksiuk and Gregory 1974; Gregory 1974) and in summer habitat (Gregory 1975; Gregory and Stewart 1975) and have concentrated on the Interlake region of the province, between Lake Manitoba and Lake Winnipeg. To date, how- ever, little work has been done on the other four species of snakes which occur in Manitoba: the western plains garter snake (Thamnophis radix haydeni), the northern red-bellied snake (Store- ria occipitomaculata occipitomaculata), the smooth green snake (Opheodrys vernalis), and the plains hognose snake (Heterodon nasicus nasicus). The subspecific status of Opheodrys vernalis in Manitoba is unclear; Grobman (1941) considers them to be vernalis X blanchardi.

Thamnophis radix, Storeria, and Opheodrys all occur in the southern part of the Interlake region, but none 1s apparently as abundant as T. sirtalis in this area. In all three cases, the southern Interlake is the northern limit of distribution for the species at that longitude (Logier and Toner 1961; Conant 1975). Tham- nophis radix, however, does range somewhat further north to the west of the Interlake region, closely following the limit, of prairie grassland. Jordan (1967) indicated that this species is

largely restricted to this habitat. Heterodon is confined to the southwestern portion of the province (Scott 1970).

While conducting field studies of T. sirtalis from fall 1969 to spring 1973, I occasionally encountered individuals of T. radix, Storeria, and Opheodrys. Since little is known about the ecology of these species, especially near the northern limits of their ranges where interesting life history strategies might be expected, I decided that it would be worthwhile to collect information on various aspects of their natural histories. Data on seasonal activity, diets, body size and age classes, and various aspects of reproduction were of particular interest as they could be readily obtained from living animals. A few attempts were made to obtain information on hognose snakes, but only one road-killed specimen was found. Specimens of Heterodon were occasionally brought in to the Department of Zoology, University of Manitoba, the most notable being a juvenile found as late as | October in 1972, but no data were obtained from any of these.

Study Area and Methods

Most of the data presented in this paper were collected from snakes found in the southern Interlake region of Manitoba. This area, des- cribed in an earlier paper (Gregory and Stewart 1975), consists of a series of ridges covered with

19

20 THE CANADIAN FIELD-NATURALIST

aspen forest and cleared farmland with marshes in the depressions between the ridges. Snakes were occasionally found in other parts of Manitoba such as around the southern part of Lake Manitoba or in the southwestern portion of the province. Both of these areas are generally more open, with considerably more grassland, than the Interlake.

No particular sampling plan was followed with respect to the collection of data as the snakes were generally found while I was search- ing for T. sirtalis. Also, because I did not decide to collect data on the three species in an organized way until I was about half-way through my studies, the data are rather incom- plete, varying from simple reported sightings to full information on particular specimens. Sampling effort was fairly evenly spread over the whole study period. In April, May, September, and October my efforts were concentrated at hibernacula of 7. sirtalis; in June, July, and August I generally searched for snakes away from known denning areas.

The time period covered by this paper is from the autumn of 1969 to the autumn of 1972. In general, captured snakes were sexed, measured (snout-vent length, SVL) to the nearest 5 mm, palpated to force regurgitation of stomach contents, and individually marked for future identification by the removal of subcaudal scutes in unique combinations. The reproductive con- dition of females was noted and some gravid females were taken back to the laboratory where further reproductive information (dates of birth, brood sizes, etc.) was collected.

Results

Thamnophis radix, Plains Garter Snake

During the course of this study, 56+ specimens of T. radix were collected in spring and fall at communal hibernacula largely occupied by T. sirtalis, more than 51 specimens were observed and/or collected at locations other than dens. Earliest and latest dates of observation were 26 April (1971) and 16 October (1970) respectively, both at den-sites. Earliest and latest observa- tions away from dens were made on 2 May (1972) and 4 October (1970).

Thirty-three of the snakes found at dens were females and 12 were males; the sexes of the

Vol. 91

remaining snakes were not noted. All but seven of the specimens from dens were caught in the fall; three of the seven found in spring were crow- kills. Only three of the snakes marked at dens were recaptured and in each case the recapture was made in the same fall season as that in which the snake was first seen; the longest interval between first and second captures was |1 days. Only one attempt was made to palpate food from a snake’s stomach at a den-site and no food was obtained.

Thamnophis radix was occasionally found at all dens numbered in Figure 2 of Gregory and Stewart (1975) and at a few others not indicated in that figure. By far the largest number (22) was observed at Den |. In addition, more specimens (22+) of T. radix were observed at dens in the fall of 1970 than at any other time. Two instances of suspect interspecific mating activity were ob- served at dens in 1972 (8 and 10 May); in both cases, a female 7. radix was being chased or mobbed by a few male T. sirtalis in the manner characteristic of mating T. sirtalis (Aleksiuk and Gregory 1974). The outcome of this activity was not observed either time as the groups rapidly dispersed upon my arrival.

Most snakes encountered away from dens in summer were found in or near marshes or in meadows adjacent to marshes; snakes were occasionally found in areas of aspen forest. Road-kills formed the majority of the more than 20 dead specimens found, while more than 31 snakes were caught or observed live. Thirty-five of the snakes sexed were females, and 13 of these were gravid. Seven of the sexed snakes were males. None of the summer-marked snakes was ever recaptured. Of 21 specimens palpated, 14 produced no stomach contents. Three snakes each contained one wood frog ( Rana sylvatica), one a chorus frog (Pseudacris triseriata), one seven unidentified tadpoles, one an unidentified leech, and the last contained unidentifiable remains.

Six of the gravid females found in summer yielded information concerning brood sizes, time of birth, and other reproductive features. These data are summarized in Table 1. The three gravid females kept in an outdoor compound at the University of Manitoba (see Table 1) also provided the only individual growth data I was

1977

TABLE 1—Summary of information on brood sizes and birth dates for Thamnophis radix

Date found S_V length,mm_ Reproductive information

22 June 690 Kept in outdoor compound

(1972) until 31 July. Gave birth to 35 live and 4 stillborn young on 3 August

22 June 710 Kept in outdoor compound

(1972) until 25 July. Gave birth to 51 live and 3 stillborn young on 8 August

3 July 645 Kept in outdoor compound

(1972) until 31 July. Gave birth to 11 live and 5 stillborn young on 24 or 25 August

3 July 580 Found as road-kill and

(1972) dissected; contained 28 embryos and 41 ovulation scars (corpora lutea)

20 July 660 Found as road-kill and

(1972) dissected; contained 14 embryos (similar to stage 37 of T. sirtalis (Zehr 1962)) and 33 ovulation scars

11 August 650 Kept in laboratory.

(1971) (10 Sept.) Gave birth to 25 live and

1 dead young between 3 and 6 September in- clusive (average SVL of 22 young was 187.3 mm)

able to obtain; these data are presented in Table ee

The SVLs of all measured snakes are shown in Figure 1; 41 of these were obtained at dens and 22 away from dens. These data are combined for all years, shown by calendar date, and divided

GREGORY: LIFE HISTORY OF THREE SNAKE SPECIES, MANITOBA 21

into males, obviously gravid females, other females, and unsexed. The lines enclosing various groups of points indicate what I feel may represent real age classes, based on apparent tendencies of increasing size throughout the summer and taking into account the size of newborn young (Table 1).

Storeria occipitomaculata, Red-Bellied Snake

No overwintering sites were known for this species in the Interlake. Most specimens were found in generally open areas such as marshes or meadows, although a few were found in areas of aspen forest. Both gray and browncolor morphs were included in the sample, but the number of each was noted. Twenty-eight specimens were found, six of them dead. Eighteen were identi- fied as females, of which eight were obviously gravid, and two were identified as males. None of those marked was ever recaptured. Earliest and latest calendar dates of observation were 31 May (1970, 1972) and 5 October (1971). Seven snakes were palpated without success for stomach contents.

No information on brood sizes was obtained, although one female produced at least two young in captivity between 28 July and 3 August in 1972. One road-killed specimen found on 6 August 1971 had six poorly developed embryos in the left oviduct. A second road-kill found on 22 August 1972 had a copulatory plug in the cloaca indicating recent mating.

Figure 2 shows the SVLs of the 12 Storeria which were measured in the same manner as 1n Figure | for 7. radix.

Opheodrys vernalis, Smooth Green Snake Forty-one green snakes were found during the

study period, 11 of them dead. These snakes

were found about equally in wooded areas and in

TABLE 2—Snout-vent lengths (mm) of three gravid female Thamnophis radix at various dates during period when held at University of Manitoba in 1972. Vertical lines indicate time interval when parturition occurred

Snake No. 22 June 3 July 25 July 9 August 29 August l 690 | 705 2 710 730 | 735 3 645 | 660

22 THE CANADIAN FIELD-NATURALIST

800

600

SVL (mm)

200

WORE EOE AS eS) July

May June

Vol. 91

29 18 7, 27 ah 3 Aug. Sept... Oct:

FIGURE |. Distribution of sizes of Thamnophis radix over active season. o = gravid female, e = non-gravid female, x = male, A = unsexed. Points between vertical solid lines represent summer-caught snakes; other points represent den-caught specimens. Enclosed groups of points represent presumed year classes, dashed lines for females and solid lines for males. Arrow connects two points representing same individual. Parenthesized numbers indicate more than

one snake of a given size at a particular date.

300

100

SVL (mm)

30

19 9 June

July

29 27

Sept.

18 7, Aug.

Zé

FIGURE 2. Distribution of sizes of Storeria over active season. Symbols as in Figure 1.

meadows adjacent to marshes. No denning sites were known for green snakes in the study area. Nineteen of the sexed individuals were females, of which eight were gravid, and 13 were males. No marked snakes were subsequently recap- tured. The earliest date of observation was 28

April (1972) and the latest 23 August (1971). Twenty-three stomach examinations were made. Seven of these produced food items. Three snakes each contained one grasshopper, two contained one caterpillar, one contained two caterpillars, and one contained arthropod

1977

remains. Further taxonomic identification of these prey items was not attempted.

Clutch sizes were obtained for four gravid females. One female measuring 285 mm SVL was found on 26 June 1972 and palpated, revealing six ova. Three females found in a group on 26 June 1972 (referred to in Gregory 1975) contained five eggs each. The SVLs of these snakes were 280, 285, and 290 mm. One of the snakes died two days after capture and was found, upon dissection, to contain eggs with shells. A second laid three eggs on 29 June and then died; the remaining two eggs had shells. The third specimen laid her eggs between 14 and 17 July, inclusive. None of the 15 eggs survived to hatching, but the conditions provided them were not optimum. A female caught on 9 July 1972 (275 mm SVL) had very loose skin in the abdominal region, similar to that of females which have recently oviposited.

Figure 3 summarizes SVL data for the 24 specimens of Opheodrys which were measured.

Discussion Habitats, Food, and Seasonal Activity

The habitats described for the three species in this paper are probably not representative since habitat type was not very often noted. In addition, most of my field time was spent in

400

ZOO

SVL (mm)

19 June

GREGORY: LIFE HISTORY OF THREE SNAKE SPECIES, MANITOBA

23

marshes and adjacent meadows, the usual summer habitat of T. sirtalis (Gregory and Stewart 1975), thus biasing the capture locations of the other three species.

Stomach examinations made of all three species suggest a similar situation to that des- cribed for T. sirtalis (Gregory and Stewart 1975): snakes feed fairly infrequently. The stomach samples obtained from T. radix, although few in number, all represent prey items also eaten by T. sirtalis. D. Platt (personal communication) found that the two species had similar diets in Kansas. No samples were obtained from Storeria, but Stebbins (1966) and Conant (1975) indicate that this species eats slugs, earthworms, and soft-bodied insects. These food types are all available in the Interlake. Opheodrys, as indi- cated elsewhere (Blanchard 1932: Judd 1960; Stebbins 1966; Conant 1975), is largely insectiv- orous.

Thamnophis_ radix, Storeria occipitoma- culata, and Opheodrys vernalis all appear to have a longer active season away from hiber- nacula each summer than do sympatric popula- tions of T. sirtalis. | have occasionally found T. sirtalis active away from den sites at the times of earliest and latest captures recorded here (the earliest capture for Storeria and latest capture for Opheodrys are ignored here), but T. sirtalis is

7

29 July

18 Aug.

9

FIGURE 3. Distribution of sizes of Opheodrys over active season. Symbols as in Figures | and 2.

24 THE CANADIAN FIELD-NATURALIST

generally congregated at denning sites at such times. It is unlikely that the early and late 7. radix, Storeria, and Opheodrys specimens were simply exceptional individuals. The relatively small number of 7: radix caught at dens in spring, as compared to fall, suggests that dispersal in this species must take place shortly after emergence. Also, those specimens found at dens in spring were generally found early in the season. Lang (1971) found that Storeria arrived later in fall and emerged earlier in spring at hibernacula in Minnesota than did either Opheodrys or T. sirtalis (mostly young). In all three species, dispersal took place very soon after spring emergence, although the overall period of emergence was fairly lengthy.

By contrast, adult male 7. sirtalis in the Interlake may spend as much as a month at den- sites before dispersing in spring (Gregory 1974), this tenure being apparently related to the intense mating activity characterizing this stage of the annual cycle. Females disperse after mating. Individual snakes of both sexes spend considerable time at the den-sites prior to entering into hibernation in the fall. For some males, the time spent in the vicinity of hiber- nacula may represent half of the annual active season.

Storeria and Opheodrys in the Interlake presumably overwinter communally in aban- doned ant mounds, as described by Lang (1969, 1971) in Minnesota and by Criddle (1937) near Treesbank, Manitoba. Being small species, they (as are young-of-year garter snakes) are prob- ably capable of using a large variety of structures in which to hibernate and likely have little difficulty locating suitable hibernacula in their summer ranges. Lang also found young 7. sirtalis in his hibernacula and Criddle found juvenile 7. radix. Young garter snakes of either species are not known normally to overwinter with the adults in the Interlake and possibly also use ant-mound hibernacula. Ant mounds of the type described by the two authors above are abundant throughout my study area but attempts to find some containing snakes failed. Adult 7. radix apparently do not usually hibernate in large groups as do adult 7: sirtalis (Gregory 1973). D. Hart (personal communi- cation), however, has observed one instance of

Vol. 91

small-scale communal denning in T. radix. In general, adult garter snakes are probably more restricted in choice of hibernacula because of their relatively large size and either find sites capable of accommodating large numbers (7. sirtalis) or hibernate in smaller numbers (T. radix).

Sex Ratios, Growth, and Reproduction

The sex ratios presented in this study are biased in favor of females and cannot be considered representative for any of the three species. This is attributable in part to the non- rigorous manner in which data were collected and to the lack of complete information for many specimens. In addition, the relative pro- portions of females for the two live-bearing species (7. radix and Storeria) are further inflated since gravid females are more easily seen and caught. This is because gravid females tend to bask in exposed areas, presumably to pro- mote development of their broods (Blanchard 1937; Gregory 1975). This does not explain, however, why more females than males of 7. radix were caught at dens. Similar samples of 7. sirtalis at the same dens generally yielded more males than females.

Thamnophis radix apparently grows at about the same rate as does T. sirtalis, females being larger than males of the same age. This is based mainly on the data presented in Figure | and in Gregory (1973). Similar growth rates are implied for 7. radix in the Chicago region (Seibert and Hagen 1947) where the growing season is from the end of May to the first week in September. This is about the same as the growing season for T. sirtalis in Manitoba. Platt (personal com- munication) found growth rates of T. sirtalis and T. radix to be similar in Kansas, where both species apparently hibernate in their summer habitat. The relatively small amount of growth shown by the captive gravid females (Table 2) may be an artifact of captivity; on the other hand, gravid females of many species of snakes do not feed very often (Gregory and Stewart 1975) so that little growth might be expected even under natural conditions.

Communal mating such as described for T. sirtalis (Aleksiuk and Gregory 1974) is not known for any of the three species studied here.

1977

There is in fact no evidence of mating activity of any kind at den sites in these three species. Possibly, however, dispersal in 7. radix, Storeria, and Opheodrys does not involve such long movements as the 4.3 to 17.7 km in T. sirtalis (Gregory and Stewart 1975) and the probability of finding a mate in the summer habitat may be greater. Lang (1969, 1971) in- dicates that Storeria move less than 3000 ft (about 914.4 m) away from their denning sites in summer.

The possibility of interbreeding between the two garter snake species is suggested by the observations of male 7. sirtalis chasing female T. radix, although I have never seen what I took to be a hybrid nor have I heard of one. Both these instances were observed near the beginning of the peak period of mating of 7. sirtalis, when the activity is extremely frenzied at den sites. Perhaps male 7. sirtalis are not very discrim- inatory at such times or the female 7. radix had been in close contact with female 7. sirta/is and retained some of their odor. Odor is very important in attracting male TJ. sirtalis to females (Fitch 1965; Aleksiuk and Gregory 1974). It is unfortunate that the end result was not observed in either case, but the lack of known hybrids suggests that nothing viable would have resulted.

Both species of garter snake normally mature in their second year (Seibert and Hagen 1947; Fitch 1965; Gregory 1973; Platt, personal communication), although Platt found that in- dividuals of either species could sometimes reproduce in their first year when food was abundant. In the Chicago area, T. radix mates in late April and early May (Cieslak 1945). Birthin T. radix in Manitoba (Figure |, Table 1) takes place from early August on, at about the same time as in 7. sirtalis (Gregory 1973). The one staged group of embryos (Table 1) is in line with development in 7. sirtalis (Gregory 1973), assuming similar stages in both species. Cieslak (1945) found that parturition of 7. radix occurred from 2-23 August, but was concen- trated in the first week of August. In Kansas, 7. sirtalis and T. radix both give birth from late July on (Platt, personal communication). The litter sizes reported for T. radix in Table | appear to be somewhat larger than for female T. sirtalis

GREGORY: LIFE HISTORY OF THREE SNAKE SPECIES, MANITOBA 25

of the same SVLs (Gregory 1973), but the sample is small. Platt indicates little difference in litter size between the two species, and Cieslak (1945) reports an average litter size in T. radix similar to that in 7. sirtalis (Fitch 1965; Gregory 1973). The average SVL of the single litter of 7. radix measured (Table 1) is towards the upper end of the range in newborn sympatric 7. sirtalis (Gregory 1973).

Accurate estimates of brood size in Storeria were not obtained, but Blanchard (1937) indi- cates a range from | to 13. My data (Figure 2) suggest that young may be born any time from early July on, but so few gravid Storeria were found that this may be very misleading. Blan- chard (1937) indicates that young are generally born from early August on, as in garter snakes. The sizes of gravid females reported here are at the low end of the range or slightly below the range reported by Blanchard, but all of my specimens were collected in early summer before much growth had taken place. Presumably, females are able to bear young by their second year (Blanchard 1937). The discovery of a copulatory plug in a female Storeria in late summer indicates that fall mating takes place, although it may simply be occasional as in 7. sirtalis (Aleksiuk and Gregory 1974). The work of Trapido (1940), however, suggests that fall mating is probably the rule and that the females retain the sperm in utero overwinter. The Storeria found in this study were comparable in size to those measured by Lang (1971).

Green snakes in Manitoba are also similar in size to those reported elsewhere (Seibert and Hagen 1947; Judd 1960; Lang 1971), females being larger than males (Seibert and Hagen 1947). They probably begin breeding in their second year (Seibert and Hagen 1947); repro- ductive females in Manitoba are about the same size as those recorded by Judd (1960). Being Oviparous, this species oviposits somewhat earlier in the summer than the other three species bear their young. The evidence is fairly strong (Figure 2, plus other observations) that eggs are laid from late June to mid-July. This is similar to the egg deposition period reported for Opheo- drys near Chicago (Stille 1954). Blanchard (1932), however, found that this species in northern Michigan laid its eggs much later than

26 THE CANADIAN FIELD-NATURALIST

this, usually in the first three weeks of August. Judd (1960) reported finding a female which had evidently oviposited very recently on7 August in southern Ontario. This disparity is interesting because Neill (1964) indicates that Opheodrys has a tendency to retain its eggs longer towards the northern part of its range, approaching a state of ovoviviparity. Stille (1954) also notes this phenomenon and Blanchard (1932) alludes to it in citing an example of a clutch which hatched four days after being laid. On this basis, one would expect a later, rather than earlier, Oviposition date for the species in Manitoba. Stille (1954), however, indicates that year-to- year variability in egg deposition dates is fairly great and that early oviposition is correlated with higher average temperatures in May, presumably because development is initiated earlier. Most of my reproductive data were collected in 1972 when May temperatures were much higher than normal (Annual Metero- logical Summary, Environment Canada, Winni- peg 1972); this possibly explains the early dates observed, but more data over several years are obviously needed. Blanchard (1932) indicates that rough handling of the females may cause premature deposition of eggs. The three females from which clutches were obtained were also palpated for food and this may have constituted sufficiently rough handling that premature ovi- position occurred. It is therefore possible that normal oviposition dates may be somewhat later.

According to Blanchard (1932), the number of eggs per clutch in smooth green snakes ranges from 3 to 11 with a mode of 7; Stille (1954) and Judd (1960) obtained similar figures. Those clutch sizes reported here are consistent with theirs. Blanchard (1932) also mentions that it is fairly common for female Opheodrys to lay their clutches in two or more batches, although this may often be associated with abnormal con- ditions. This may have been the case with the female that laid three eggs and died before laying the last two.

Conclusion

Evidently, on the basis of the limited data presented here, the ecology of 7. radix, Storeria, and Opheodrys in the Interlake region of

Vol. 91

Manitoba is not greatly different from else- where in their ranges. This is interesting since Interlake populations of 7. sirtalis exhibit many life history features that are considerably dif- ferent from those of populations in other parts of the range (Aleksiuk and Gregory 1974; Gregory 1973, 1974; Gregory and Stewart 1975). Tham- nophis sirtalis is apparently much more abun- dant in my study area than the three species reported on here and perhaps these life history features represent major adaptations which at least partially explain its numerical superiority (Gregory 1973, 1974). The Interlake environ- ment, however, is extremely rigorous for a reptile (Gregory 1973) and further study of the other three species may reveal interesting eco- logical and/or physiological adaptations. A comparison of the life histories of 7. radix and T. sirtalis would be of particular interest since parameters such as annual growth and repro- ductive potential appear to be similar in the two species in Manitoba despite obvious differences in their annual activity cycles. In other areas where the two species occur, they are apparently similar in both respects.

Acknowledgments

I thank Don Hart (University of Manitoba) for his invaluable field assistance. Both he and Dr. Dwight Platt (Bethel College, North New- ton, Kansas) provided me with valuable per- sonal communications. Two reviewers also made useful comments on the original manu- script. Jeff Lang (University of North Dakota) kindly allowed me access to his M.Sc. thesis.

Literature Cited

Aleksiuk, M. and P.T. Gregory. 1974. Regulation of seasonal mating behavior in Thamnophis sirtalis parie- talis. Copeia 1974: 681-689.

Blanchard, F. N. 1932. Eggs and young of the smooth green snake, Liopeltis vernalis (Harlan). Papers of Michigan Academy of Science, Arts and Letters 17: 493-508.

Blanchard, F. N. 1937. Data on the natural history of the red-bellied snake, Storeria occipitomaculata, in Michigan. Copeia 1937: 151-162.

Cieslak, E.S. 1945. Relations between the reproductive cycle and the pituitary gland in the snake Thamnophis radix. Physiological Zoology 18: 299-329.

Conant, R. 1975. A field guide to reptiles and amphibians of eastern and central North America. 2nd ed. Houghton Mifflin. xvin + 429 pp.

1977

Criddle, R.S. 1937. Snakes from an ant hill. Copeia 1937: 142.

Fitch, H.S. 1965. An ecological study of the garter snake Thamnophis sirtalis. University of Kansas Publications, Museum of Natural History 15: 493-564.

Gregory, P. T. 1973. Life history parameters of a popula- tion of red-sided garter snakes (Thamnophis sirtalis parietalis) adapted to a rigorous and fluctuating environ- ment. Ph.D. thesis, University of Manitoba, Winnipeg, Manitoba. viii + 98 pp.

Gregory, P. T. 1974. Patterns of spring emergence of the red-sided garter snake (Thamnophis sirtalis parietalis) in the Interlake region of Manitoba. Canadian Journal of Zoology 52: 1063-1069.

Gregory, P. T. 1975. Aggregations of gravid snakes in Manitoba, Canada. Copeia 1975: 185-186.

Gregory, P.T and K.W. Stewart. 1975. Long-distance dispersal and feeding strategy of the red-sided garter snake (Thamnophis sirtalis parietalis) in the Interlake of Manitoba. Canadian Journal of Zoology 53: 238-245.

Grobman, A.B. 1941. A contribution to the knowledge of variation in Opheodrys vernalis (Harlan) with the descrip- tion of a new subspecies. Museum of Zoology, University of Michigan, Miscellaneous Publications 50: 1-38.

Jordan, O. R. 1967. The occurrence of Thamnophis sirtalis and Thamnophis radix in the prairie-forest ecotone west of Itasca State Park, Minnesota. Herpetologica 23: 303-308.

Judd, W. W. 1960. Observations on the habitat, food, reproductive state and intestinal parasites of the smooth green snake at London, Ontario. Canadian Field- Naturalist 74: 100-106.

GREGORY: LIFE HISTORY OF THREE SNAKE SPECIES, MANITOBA 27

Lang, J. W. 1969. Hibernation and seasonal movements of Storeria occipitomaculata in northern Minnesota. (Ab- stract.) Journal of Herpetology 3: 196-197.

Lang, J. W. 1971. Overwintering of three species of snakes in northwestern Minnesota. M.Sc. thesis, University of North Dakota, Grand Forks, North Dakota. xi + 97 pp.

Logier, E. B.S. and G. C. Toner. 1961. Check list of the amphibians and reptiles of Canada and Alaska. Royal Ontario Museum, Life Sciences Division, Contribution Number 53. 92 pp.

Neill, W. T. 1964. Viviparity in snakes: some ecological and zoogeographical considerations. American Naturalist 98: 35-55.

Scott, V. H. 1970. The western hognose snake in Mani- toba. Zoolog 11(1): 15-19.

Seibert, H.C. and C. W. Hagen, Jr. 1947. Studies on a population of snakes in Illinois. Copeia 1947: 6-22.

Stebbins, R.C. 1966. A field guide to western reptiles and amphibians. Houghton Mifflin. xiv + 224 pp.

Stille, W. T. 1954. Observations on the reproduction and distribution of the green snake, Opheodrys vernalis (Harlan). Natural History Miscellanea, Chicago Academy of Science 127: I-11.

Trapido, H. 1940. Mating time and sperm viability in Storeria. Copeia 1940: 107-109.

Zehr, D. R. 1962. Stages in the normal development of the common garter snake, Thamnophis sirtalis sirtalis. Copeia 1962(2): 322-329.

Received 7 May 1976 Accepted 28 September 1976

Status and Habits of the Cougar in Manitoba

ROBERT W. NERO! and ROBERT E. WRIGLEY?

\Manitoba Department of Renewable Resources and Transportation Services, 1495 St. James Street, Winnipeg, Manitoba R3H 0W9

2Manitoba Museum of Man and Nature, 190 Rupert Avenue, Winnipeg, Manitoba R3B ON2 Nero, Robert W. and Robert E. Wrigley. 1977. 91(1): 28-40.

Abstract. A cougar (Felis concolor missoulensis), collected at Stead, Manitoba in 1973, and 281 well documented sightings establish the species as resident in the province for the first time. Though there are cougar sightings from 1879 to 1975, the majority are recent (40 in 1974). Prior to 1940, the cougar was restricted to the grassland and aspen-oak transition of extreme southwestern Manitoba—within the ranges of mule deer and American elk. After 1940, these two prey species became rare and localized, and white-tailed deer became the dominant big-game animal, spreading far north into the boreal forest. Recent cougar distribution is closely associated with that of the white-tailed deer, an apparent doubling of the cougar’s range. A rough estimate of the cougar population in Manitoba is 50, and it seems likely that some individuals have been observed in

Status and habits of the cougar in Manitoba. Canadian Field-Naturalist

adjacent Saskatchewan, North Dakota, Minnesota, and Ontario.

Until recently, it was generally assumed that the cougar had been exterminated east of the Rocky Mountains, except for a small Florida population of between SO and 100 animals (Wright 1972). Occasional sightings continued, however, in every Canadian province (e.g., Saskatchewan, White 1967; New Brunswick, Wright 1959, 1972) except Prince Edward Island and Newfoundland, and in many central and eastern states (e.g., Pennsylvania, Doutt 1969; Maine to Alabama, Wright 1972). Most reports (from non-biologists) were not taken seriously, and consequently, valuable information that should have been accumulating on this en- dangered species has been lost.

The objectives of this paper are to describe the history and present status of the cougar in Manitoba (and to a lesser extent in adjacent regions), relate its distribution in the province to prey and cover, and examine the animal’s habits in this part of its range.

Early historical accounts offer little informa- tion on the presence of cougars in Manitoba. The fur trader, Alexander Henry, made no mention of this cat in his extensive travels throughout the southern part of the province and adjacent states during the years 1799 to 1808 (Coues 1897). A possible reference to cougars was the name, Tiger Hills, given by early settlers to a plateau situated south of the Assiniboine River between the Cypress River, Wawanesa, and Ninette.

The first accounts of Manitoba mammals by Thompson [Seton] (1886) and Seton (1909) did not include the cougar, but ina subsequent work Seton (1925-1928) listed seven localities in southwestern Manitoba where the animal had been seen or shot. A few additional records found their way into the literature, but despite the availability of a specimen taken just over the border in Saskatchewan in 1948 (Beck 1958), the absence of any authenticated Manitoba speci- men in a museum resulted in this province being omitted from northern distribution limits plot- ted for the species by Young and Goldman (1946) and by Hall and Kelson (1959).

In 1973 a cougar was shot in a farmyard at Stead, 56 km northeast of Winnipeg. The speci- men, which is now in the Manitoba Museum of Man and Nature, gives substantial support to the mounting evidence that Manitoba and ad- jacent regions of midwestern Canada and the United States support a resident cougar popula- tion, almost 1000 km east of the recently accepted range for the species.

Methods

Most of the early records from Manitoba were published by Seton (1925). Thereafter, reports were investigated and recorded by two former directors of the Manitoba Museum, L.T. S. Norris-Elye (1951) and his successor, Richard W. Sutton (1960). W. Harvey Beck, an assistant curator at the Manitoba Museum of Man and

28

Oa

Nature, started a card file on cougar records in 1968 to which were added records gathered by Charles H. Buckner, Federal Forest Biology Laboratory in Winnipeg. In 1972 the senior author initiated a program to obtain and record additional past and new cougar observations in order to document further the status of this species in Manitoba. At that time the cougar was not identified as a game species or as a predator and, if anything, was regarded as a rare transient. Letters and a preliminary report were sent to field staff of the Department of Re- newable Resources and Transportation Ser- vices, requesting their cooperation in the project, and a questionnaire was sent to Provincial Game and Fish associations. No attempt was made to advertise broadly our interest in receiving further reports from the general public. Many reports were investigated further by a letter, telephone call, or personal contact with the original observer, in the hope of obtaining date, location, time, lighting conditions, length of observation, distance to animal, color, descrip- tion, behavior, reasons for suspecting cougar, and number of observers. Most observers indicated a reluctance to publicize their obser- vation for fear of inviting ridicule, and it was learned that a number of sightings had been reported to local authorities at an earlier date, but went unrecorded before our survey owing to a general skepticism regarding the occurrence of the cougar in Manitoba.

Finally, cougar data were requested from wildlife agencies, museums, and universities in Saskatchewan, Ontario, North Dakota, and Minnesota to compare trends of sightings and the extent of adjacent cougar populations.

This study is based on observations by people with a wide range of occupations and back- grounds. How credible are these sightings? Obviously, there is no way to prove the reli- ability of a report short of obtaining a specimen or photograph. Most reports accepted by us as valid and used in this study consisted of a good description of the animal, including its behavior, and pertinent details regarding the circum- stances of the sighting. Our judgment was also based on other factors; for example, when a number of people saw cougars in the same region, unknown to each other, the evidence became more convincing. Other kinds of evi-

NERO and WRIGLEY: THE COUGAR IN MANITOBA 29

dence such as tracks, calls, and the presence of partly eaten wild or domestic prey, provided support for some records, but otherwise were not used to establish the occurrence of the species. Knowing that certain other species may be confused with the cougar, namely, deer, wolf, coyote, dog, and lynx, particularly if the obser- vation is made at a great distance or under poor conditions, we made a special effort personally to check reports. Many sightings were rejected on the basis of insufficient evidence.

Results Cougar Sightings

There are now records of 281 sightings of cougars in Manitoba for the period from 1879 to 1975, four prior to 1900, and including three in which the animal was allegedly killed (Table 1). From 1900 to 1950 there are one to seven reports per decade; however, 10 cougars were supposed- ly killed during those 50 years (Figure 1). Un- fortunately none of these specimens was saved (several were exhibited for a time in local store windows), partly because the provincial museum was not established until 1932. Since 1950 there has been a great increase in the number of reports received, with 159 during the 1970s and 40 for 1974 alone. Numerous sightings have

TABLE 1—Number of cougars seen and reported killed in Manitoba from 1879 to 1975

Year Number of Number of sightings kills 1870-1879 | 1 1880-1889 l ] 1890-1899 2 | 1900-1909 7 4 1910-1919 1 | 1920-1929 5 3 1930-1939 6 2 1940-1949 4 0 1950-1959 27 0 1960-1969 68 ] 1970 8 0 1971 14 0 1972 37 0 1973 35 1 1974 40 0 1975 25 0 Totals 281 15

30 THE CANADIAN FIELD-NATURALIST

undoubtedly gone unreported from earlier years.

Many observers supplied surprisingly good descriptions of the cougar. The following two reports, for example, are typical of many in our files: Murray Thompson, describing an animal he observed in Riding Mountain National Park in midafternoon in August 1972, wrote as follows: “The animal appeared to be tawny brown or light brown, about 4.5 to 5 feet [1.4-1.5 m] long, with a long, long tail like a piece of heavy rope. It had heavy limbs and large paddy feet, and its head appeared about two sizes too small for its body. The animal was seen walking down the road, then it crossed a ditch and disappeared into the bush. It seemed to pay little attention to the car.”

In September 1972, RCMP Constable John Ireland noticed a road-killed deer in the ditch while he was on patrol on Highway 59 at Birds Hill Park, just 13 km northeast of Winnipeg. Returning at 0200 hours, he observed a cougar standing over the deer only 5 m away — soclose that the whiskers on the cat’s upper lip were clearly visible. Constable Ireland guessed the weight of the cougar to be about 35 kg. The cat looked into the headlights for several minutes before it ran a few metres, then came back and tried to drag the deer away. After two attempts failed, it ran off into the bush. The following morning the deer was found where it had been dragged 70 m into the bush; it was partly eaten around the rib cage and was covered with leaves — typical cougar signs. By the next day the deer had been all eaten except the vertebral column, head, legs, and skin.

Cougars Killed in Manitoba

On the night of 25 December 1973, three men investigated a commotion caused by a barking dog and disturbed cattle in the farmyard of William Kowalke at Stead, 56 km northeast of Winnipeg. A large animal was spotted with the flashlight and, believing it to be a wolf intent on the livestock, the men shot what turned out to be an adult cougar. The specimen was eventually obtained by the Manitoba Museum of Man and Nature (Cat. No. 4725), and is the first con- firmed record of the cougar in Manitoba. The animal was a thin, 2-year-old male (aged by M. G. Hornocker, University of Idaho), measur-

Vol. 91

ing as follows: total length, 2108 mm; tail length, 800 mm; hind foot, 292 mm; ear, 95 mm; testes length, 25 mm; weight, 43 kg. Male cougars (F. c. missoulensis) from western Canada aver- age 73 kg and 2413 mm in total length (Cowan and Guiguet 1956).

There may have been at least 14 other cougars killed in Manitoba since 1879 (Table 1), but none has been substantiated by a museum specimen. A mounted cougar in the J. B. Hales Museum in Brandon may be the animal that was shot 19 km southeast of Brandon in the early 1920s, but this has not yet been proved. The following cougar kills are listed according to date: Pembina Hills (1879); “Plum Coulee in the Souris Country” [probably Plum Creek] (1887); Birtle (about 1895); Makinak (1901-1902); Elphinstone (1904); Brandon (two killed in 1904, skins on exhibit in general store but now lost); Duck Mountain (“early days”); Little Souris River, 19 km SE Brandon (early 1920s); Birtle (1922); Alexander (1926); Pendennis, 15 kmN, 16 km W Brandon (early 1930s, skin exhibited in general store); Souris River Bend area, about 6 km N, 3 km E Margaret (1936-1937); 3 km N, 15 km W Hadashville (1969); 15 km S, 5 km E Stead (1973).

Season and Time of Sightings

Sightings were recorded from every month, and 220 could be assigned to particular seasons as follows: 40%, summer (June to August); 33%, autumn (September to November); 14%, winter (December to February); 13%, spring (March to May). These results are expected, since summer and autumn are the seasons when people spend the most time travelling on back roads and working on the land. In New Brunswick, Wright (1959) also reported sightings during every month, with 71% observed from June to November, compared to 73% in this study.

The exact time of sighting was recorded in 86 cases. Cougars were seen during every hour of the day and night; from 2100 hours to 0600 hours (sun generally at or below the horizon) there were only one to four reports each hour, whereas the peak periods were 1100 hours and 1600 hours (nine reports each) and 0800 hours (seven). There were 26 (32%) reports from 0700 to 1100 hours and 42 (50%) from 1400 to 2000 hours. Cougar activity periods vary according to prey

1977 NERO and WRIGLEY: THE COUGAR IN MANITOBA 31

ie eee pa

HS Island Lak}

Baie ey Ing

e Swan River

» Winnipegosis DUCK MIN

a fy

e Russell RIDING TN e@ McCreary

Shoal Lake

| %

Selkirke

ae aN Winnipeg

Steinbach

S \ ® Morris x

Killarne —~e Morden L

Emerson

FIGURE |. Map of southern Manitoba showing the maximum distribution of mule deer and American elk (solid line) (after Seton 1909), and cougar sightings (dots) and reported kills (stars) prior to 1940.

32 THE CANADIAN FIELD-NATURALIST

activity, and although most hunting is done at night, cougars have also been known to kill big game in the evening and midmorning (Seiden- sticker et al. 1973).

Sightings of Pairs and Young

The cougar may give birth to from one to six cubs at any time of the year, but breeds only once every two or three years. Young remain with the mother until almost full grown at about 2 years of age. Adults are typically solitary and avoid each other, except for short periods of a few weeks when two females may associate or whena male may be courting a female in heat (Seiden- sticker et al. 1973).

In Manitoba, sightings of two or more animals at the same time were reported on 14 occasions. Four cases involved pairs of adults, or adults and large young not distinguished as such; three cases of adult(s) and young (two adults and one young, two adults and five young, one adult and one young); three cases of an adult with cubs (twice of one adult and two cubs, one adult with one cub); two cases of two young alone; two cases of cubs alone (four cubs, two cubs); and two instances when single young cougars were observed, one of which was treed by dogs and then shot. Some of these observations lasted many minutes, with cougars running or walking across open fields, jumping over fences, or climbing into trees. Three reports of paired adult(s) and/or young were made independently by different people in the same vicinity and season.

Vocalization

In the present investigation, loud cougar calls were described in 15 reports, including four while the animal was in full view, and a cougar was seen in the area in eight of the remaining 11 cases.

A difference of opinion has existed among cougar experts as to whether the cougar emits a loud scream. Some observers (e.g., Seiden- sticker et al. 1973) admit hearing close-range vocalizations, but never roaring and screaming such as is abundantly described in the literature (Young and Goldman 1946; Barnes 1960). Although calls of bobcats (Lynx rufus) and owls have probably been ascribed incorrectly to cougars On some occasions, there is no doubt

Vol. 91

that cougars are capable of producing terrifying calls, since they have been observed screaming in the wild and in captivity. It appears that cougars are rather silent except at mating time (Seton 1925; Wright 1959), and this may partly account for the fact that cougars are heard in one region but not in another.

Behavior

A peculiar aspect of cougar behavior, pointed out repeatedly in these reports as well as in the literature, is its indifference to being observed from a distance by people on foot or in motor vehicles. Cougars often permitted the approach of a car to within 10 to 100 m before walking leisurely or bounding to the nearest cover. The following case illustrates how bold the cougar may be. On 30 June 1974, at 1400 hours, Tim Sims was cultivating a field near Snowflake, Manitoba, when a cougar came out 400 m from the bush and lay down in a fresh furrow, per- haps to escape flies. Sims watched the animal for an hour as he made several rounds of the field, once passing within 12 m (close enough to see the cougar curl its lips and snarl). It did not appear to be frightened. Sims had seen the same animal in this field 3 days earlier and described it as brown, 0.8 m in height, 2.7 m from nose to tip of tail, tail about 0.6 m with a dark tip. On 8 August Mr. Sim’s son saw a cougar with three cubs on the same land.

Food Habits

Cougars were seen stalking, chasing, or eating a variety of prey species, and were implicated (by nearby sightings, screams, tracks) in many others. Attacks (prey killed, wounded, or flee- ing) were reported on cattle (13 cases), white- tailed deer (Odocoileus virginianus) (9), horse (3), snowshoe hare (Lepus americanus) (3), chicken and turkey (2), Mallard (Anas platyr- hynchos) (1), hog (2), sheep (1), moose (Alces alces) (1), Formosan deer (1), dog (2), fisher (Martes pennanti) (1), and man (1 possible fatality). Cougars were seen stalking the follow- ing: cattle (3 cases), white-tailed deer (3), beaver (Castor canadensis) (1), pheasants (Phasianus colchicus) (1), and ravens (Corvus corax) sca- venging on a muskrat (Ondatra zibethicus) carcass (1). Cougars were seen scavenging on a white-tailed deer carcass, a trapper’s cache of

1977

carcasses, a trapline (tracks and other sign indicated scavenging ona snared snowshoe hare, and killing and eating a fisher caught in a trap), and in a refuse dump. There were three instances of cougars observed drinking.

The Stead cougar had striped skunk ( Mephitis mephitis) hairs in the digestive tract, and porcupine (Erethizon dorsatum) quills in the muscles of the forearms and shoulder regions. Traces of both species are often found in scats of cougars, and porcupines appear to be favored prey (Young and Goldman 1946; Wright 1959; Barnes 1960).

One report is of exceptional interest. In the summer of 1972, Walter Larocque herded 32 horses into an old pasture surrounded by bush at The Pas. In an abandoned house on the site, he found moose droppings all over the floor and blood on the door jamb, and concluded the moose had been chased into the house by some animal. One month later he noticed that two horses were missing, and the depredations con- tinued until a total of six mares and one colt had been killed and eaten. The carcasses were found in the bush with all the meat stripped from the bones. It appeared that attacks were launched from a big elm tree that leaned over a path taken by the horses. The horses were herded into a wire corral and when two men went to check them the following morning, they had broken out. The men went to the northend of the pasture and met the horses on the run, followed closely by two tan-colored cougars which came to a halt, their “tongues sticking out of their mouths” from exhaustion of the chase. The larger cat crouched and crawled within 30 m before the men backed away. The horses were gathered and removed from the pasture. Horses, particularly colts, appear to be a favorite prey of the cougar; the large, nearby race of cougar formerly inhabiting North Dakota, F.c. hippolestes, was named for this reputation as a “horse killer.”

Because many livestock losses go unreported and government field staff presently lack ex- perience in identifying cougar kills or damage, there may be a larger incidence of cougar preda- tion on livestock than is presently known.

There is a single reported case of a cougar stalking people in Manitoba, which ended in tragedy. The following account was published by V. W. Jackson (a University of Manitoba

NERO and WRIGLEY: THE COUGAR IN MANITOBA 33

zoologist) in the Winnipeg Tribune (14 February 1942): “Twenty years ago [1922]a mountain lion attacked a boy and a girl at Birtle, Man., killing the latter. When shot it was found to be blind and half-starved.” A search for additional infor- mation on this case, including a local history of the area (Abra 1974), produced no substantiat- ing evidence.

Habitat and Distribution of the Cougar and its Prey in Manitoba

Remarks on habitat in 255 reports were grouped as follows: 40% “wilderness” (true wilderness and heavily wooded regions with few roads or human habitations), 30% “mixed land” (agricultural lands and towns interspersed with large tracts of forest), and 30% “farmland” (cropland and pasture with forest cover re- stricted to woodlots and river valleys).

The pre-1940 records (Figure 1) cover an area of about 87 000 km? and are restricted to the grassland region and the aspen-oak transition (now largely farmland and mixed land) in extreme southwestern Manitoba. From 1941 to 1975 the cougar appears to have extended its range (Figure 2) northward into the boreal forest region and eastward into the Great Lakes - St. Lawrence forest transition (Figure 3). The species now inhabits an area of about 200 000 km? in the southern half of the province, more than twice its apparent former range.

Although the cougar accepts an extremely wide variety of prey species (e.g., mice, fish, grasshoppers, small birds, etc.), its residence in an area is still dependent on big game, partic- ularly deer (Barnes 1960). “The range of the panther [cougar] has always coincided with that of the deer, and as the deer have flowed back over the eastern ranges, so has the panther returned to many of its old haunts” (Wright ISS),

In early historic times southern Manitoba supported abundant mule deer (Odocoileus hemionus), American elk (Cervus elaphus), moose, caribou ( Rangifer tarandus), pronghorn (Antilocapra americana), and American bison (Bison bison). Rather suprisingly, of these species, only mule deer and elk have been found to play a role in the cougar’s diet. Reported attacks or predation on the other four species are completely absent or restricted to a few isolated

34 THE CANADIAN FIELD-NATURALIST Vol. 91

(>) a)” | }.Nelson House A, j ‘Thompson. a7} j

',

ah) ee J

oi Za ipinesl

Lake Lye y, af Zz

G

p Cormorant

Lake* % t Jor sale Vie i)

AD) OO oe

TheyPas)& iD Fe Ctr e P

i

© Swan River

| DUCK MIN

!

e Russell VG ae N Q McCreary @® @

Neepawa Selkirkef@ e oe e > Portage la Prairie <A $

/ < @ Brando AO / e y ee Wi, e Carberry r) e Steinbach ®

é e\e ee = \ @ \ ) e Morris

& Killarne a ‘ wwe y Morden J -) e TURTLE 0 Pen ; ye ill Emerson {a ~ = ° 20

ee

60 =!

— KILOMETRES

obo

FIGURE 2. Map of southern Manitoba showing the maximum distribution of white-tailed deer (line) and cougar sightings (dots) from 1941 to 1975. The hollow star indicates the location of the 1969 cougar kill, and the solid star, the Stead specimen killed in 1973.

1977

observations (Young and Goldman 1946; Barnes 1960).

Mule deer and elk recovered from overex- ploitation of the 1870s and 1880s in Manitoba, and both became numerous throughout the grassland and aspen-oak transition at the turn of the century (Seton 1909). By the end of the 1930s, however, both species were drastically reduced to relatively isolated populations, mostly in the Interlake region between Lakes Manitoba and Winnipeg, the rugged country of the Manitoba Escarpment (Pembina Hills; Turtle, Riding, Duck, and Porcupine Moun- tains), and the Carberry Sandhills. Probably asa result of competition with white-tailed deer, habitat destruction, and hunting, the mule deer has now been all but extirpated in Manitoba.

At present there are six herds of elk in Mani- toba, totalling 5540, of which about 55% (2300-3000 animals) inhabit Riding Mountain. The Duck Mountain herd (1500-2000 animals) and the Porcupine Mountain herd (125-150) often move back and forth from adjacent rugged hill country in Saskatchewan. The Carberry Sandhills herd numbers about 150; the Man- tagao herd west of Hodgson, up to 160; and the northern Interlake herd northwest of Gypsum- ville, 80. All but the latter two areas are regions of concentrated cougar sightings. Figure | illustrates the maximum ranges of mule deer and elk in Manitoba (Seton 1909). Cougar sightings from the late 1800s to 1940 are confined to the former ranges of these big-game prey species.

The white-tailed deer invaded Manitoba from the Red River Valley of Minnesota and North Dakota around 1881 and became common in many areas by the early 1900s (Seton 1909; Criddle 1929). By the 1940s this species had completely replaced the mule deer as the common deer of the province. According to pro- vincial wildlife biologists, light hunting pressure and ideal habitat conditions of the 1940s allowed deer populations to peak in the early 1950s, perhaps numbering close to one-quarter million animals. Typical of most species invading new range, their numbers soon leveled off and then began to recede. Continued clearing of forest for farmland, severe winter climate, and heavy hunting pressure caused a steady decline through the late 1960s, culminating in the brutal winter of 1973-1974 which left a remnant herd of

NERO and WRIGLEY: THE COUGAR IN MANITOBA 35

less than 50 000 animals.

Figure 2 shows the white-tailed deer’s maxi- mum distribution (far into the boreal forest), and sight records of cougars from 1941 to 1975. About 85 cougar reports are outside the cougar’s former range and the pre-1940 ranges of mule deer and elk, and closely approximate the distribution of white-tailed deer. North of The Pas and Grand Rapids, deer are scarce and localized. The poor correlation in eastern Mani- toba north of Bissett is probably due to the absence of roads (and hence observers) in this wilderness region, as well as to the scarcity of deer and cougar along this northern boundary of their range. In interpreting the correlations of cougar distribution with that of mule deer, elk, and white-tailed deer, it should be kept in mind that population estimates of these prey species were not well known until recently, and that there are presently many more observers in the northern parts of these ranges than in the past.

The distribution of cougar records is closely associated with hilly country (e.g., Riding Mountain) and well-forested river valleys (e.g., Assiniboine River and its tributaries), especially in the grassland region. Sightings appear to be especially numerous over the years in certain locations — the region between Swan River and Duck Mountain, Winnipegosis, Brandon, the Pine Dock road along the west shore of Lake Winnipeg, Bissett, the Winnipeg River, and Sprague — most of which offer dense forest cover and concentrations of white-tailed deer.

Status of the Cougar in Adjacent Provinces and States

Transient cougars have been known to make incredibly long journeys in search of new range. Percy and Penelope Dewar reported (in Gregg 1974) one travelling 64 kmin one night in British Columbia. Cougars are therefore effective col- onizers, but will not remain in an area for long without other cougars nearby, even though each individual is solitary (Seidensticker et al. 1973). These facts may explain why cougars are occasionally observed by a number of people in an area far from known cougar range, and then are never seen again.

Resident cougars also travel over extensive areas. Seidensticker et al. (1973) followed five cougars over a full year in Idaho and found

36 THE CANADIAN FIELD-NATURALIST

home ranges were from 173 to 453 km2. The vegetation-terrain/ prey abundance-vulnerabil- ity complex was important in determining the range of a resident’s movements through the seasons.

Since many of Manitoba’s cougar sightings are within the animal’s range capability of Saskatchewan, North Dakota, Minnesota, and Ontario, we examined cougar records in these adjacent regions. The following reports, which

Ds BRANDON

Vol. 91

include published and unpublished accounts, probably do not present a complete picture of the species’ occurrence in the Midwest (Figure 3).

Saskatchewan

There are four records of cougars killed (authenticated by specimens) during the 1900s in Saskatchewan: Fort Walsh in the Cypress Hills, 1912 (Soper 1961); Kindersley, 1939 (Clarke

wATCOMB

IE MINNESOTA,

FiGURE 3. Map illustrating the recent distribution of the cougar in Manitoba (hatched area), cougar reports from Saskatchewan, North Dakota, Minnesota, and Ontario (dots), and the major vegetation zones of this region.

1977

1942): Connell Creek in the Pasquia Hills, 1948 (Beck 1958); and Cutknife, 17 November 1975 (56-kg female, specimen now in the Saskatch- ewan Museum of Natural History) (personal communication, F. W. Lahrman). White (1963) believed that Saskatchewan afforded suitable conditions for cougars in two areas, the semi- arid grassland and forest of the Cypress Hills in the southwest, and the heavily forested region of the Pasquia Hills in east-central Saskatchewan adjacent to the Manitoba border. He recorded more than 60 sightings in the latter region, and in a Subsequent paper (1967) described at least two kills in these hills, including the 1948 record of a cougar shot while caught in a wolf trap. Riome (1973) recorded an additional 10 sightings in the Nipawin-Pasquia Hills area from 1934 to 1973. These reports tie in to numerous records within 160 km of the elevated escarpment and valleys of adjacent Manitoba.

Further south are records at Regina (Brazier 1960), Qu’Appelle Valley north of Wolseley (White 1967), and four separate sightings includ- ing an adult and cubs, between 1970 and 1972 at Antler on the Saskatchewan border (White 1973). These latter sightings were within 16 km of six sightings in Manitoba during the years 1966, 1967, 1970, 1972-1974, and probably represent the same animals.

North Dakota

C.R. Grondahl (personal communication 1974), of the North Dakota Game and Fish Department, mentioned that an average of six reports of cougars are received per year, mostly from the western and northwestern parts of the State, remote areas with good populations of mule and white-tailed deer as well as domestic cattle.

The most recent authenticated cougar kills were reported by Bach (1943) in 1902, 40 km down the Missouri River from Williston: one was shot, and a second adult and a subadult were trapped. In the northern half of the state, files of the Game and Fish Department record sightings at Watford City (1965), Pick City (1959), Velva (1962), two in the Fort Totten Indian Reserva- tion (both in 1959), and three in nearby areas — Maddock (1959), Grace City (1959), and Cooperstown (1960). One final sighting came from Walhalla (1969), reported by Paul Crary

NERO and WRIGLEY: THE COUGAR IN MANITOBA 37

(personal communication 1974). These records are within 8 to 190 km of the Manitoba border.

Minnesota

No cougar is known to have been killed in Minnesota in the 1900s, but Bue and Stenlund (1952) stated that there are many sightings which seem authentic. “The observations came from two general regions, the northeastern part of the state [heavily forested], and agricultural southwestern Minnesota.” These northeastern sightings plus additional ones listed by William H. Longley (personal communication 1974 and 1975, including list supplied by Patrick Karns in 1964) of the Minnesota Department of Natural Resources, may be summarized as follows: one sighting in Koochiching County (1950), 17 in St. Louis County (1945-1974), nine in Lake County (1948-1962). These records come from the hilly and well forested region between Lake Superior and Manitoba, and range within 190 to 360 km of the border. About 200 km south of Sprague, Manitoba are three records from central Min- nesota: Cass Lake, Cass County (1950); Itasca State Park, Clearwater County (1971); and Many Point Lake, Becker County (1956).

Closer to Manitoba are these: a sighting at Williams in 1954 (Magnus 1956), and two 1974 sightings at Salol and Northwest Angle (W. S. Adams, personal communication 1974). The last two records are both within 40 km of Sprague, Manitoba, and quite likely represent the same individual(s) seen around Sprague on six oc- casions from 1973 to 1975.

Ontario

No cougar is known to have been killed in Ontario during the present century, but there are numerous sightings during the last 22 years in northwestern Ontario, just north of Lake Supe- rior. Dear (1955) mentions four records in the Thunder Bay area during 1953-1954. In 1966, C. H. D. Clarke, Chief of the Ontario Fish and Wildlife Branch, was quoted by Wright (1972, p. 96): “The other area [in Ontario] for which I give credence to reports [of cougars] is on the Manitoba boundary.”

Most of the following records from Ontario were supplied by Ken J. Chambers of the Ontario Ministry of Natural Resources. In the Kenora District are sightings at Camp Robinson

38 THE CANADIAN FIELD-NATURALIST

(1959), Contact Bay south of Dryden (1975), four places near Watcomb (1966-1971), and 10 around Kenora (1961-1974). These Kenora-area reports are within 16 to 70 km of Manitoba, and likely represent the same cougar population occupying the Winnipeg River drainage and, in particular, the Whiteshell Provincial Park region of Manitoba where there are numerous sightings. The Stead, Manitoba cougar was killed only 82 km from the Ontario border.

Farther south, in the Rainy River District, are the following: a single 1965 record at Atikokan and a cluster of five others between Fort Frances and Lake of the Woods (1966-1972), 80 to 130 km from Manitoba and just across the inter- national boundary from the numerous records in Minnesota.

Geographic Affinities of the Stead Cougar

Though one specimen is insufficient to make a definite subspecific determination of the local populations, it does shed some light on the geographic affinities of Manitoba cougars. Four races have been described from regions adjacent to the province. Felis concolor couguar occupied the northeastern quarter of the range in North America (from Wisconsin eastward) and was long thought to be extinct. But, it has survived in New Brunswick and possibly in other areas as well, and appears to be reinvading former territory (Wright 1972).

An extinct race, F. c. schorgeri, was described from central North America (Kansas to Min- nesota) on the basis of a 100-year-old mounted specimen and two other skulls. The lack of specimens has prevented a satisfactory study of this race’s taxonomic features or its former range. Recent surveys from the region have not used the name schorgeri for the race formerly occurring there (Jones 1964 (Nebraska)), or have accepted it with strong reservations (Bowles 1975 (Iowa)).

Felis c. hippolestes occurs to the southwest of Manitoba as far as the Rocky Mountains of Utah and Colorado. The closest specimen to Manitoba taken recently was one in the Black Hills of South Dakota in 1958, where the species remains an uncommon inhabitant of this rugged and remote area (Turner 1974). Felis c. mis- soulensis 1s a western race ranging from British Columbia to Saskatchewan and adjacent states.

Vol. 91

The Stead specimen was sent to the Mammal Section of the National Museum of Natural History, Washington, D.C., and Wildlife Biolo- gist A. L. Gardner returned the following identi- fication: “We have examined your specimen of Felis concolor and have directly compared it with specimens of F. concolor missoulensis, F. concolor hippolestes, and F. concolor couguar. In our opinion, your specimen is much closer to F. concolor missoulensis than it 1s to any of the other named forms. The broadly flaring zygo- matic arches, the large auditory bullae, large skull, and the color pattern of the skin are characters that are found in F. c. missoulensis but are not matched in this combination in the other named subspecies.” Whether this race originally inhabited or recently invaded Mani- toba is of course not known without earlier comparative specimens.

Discussion

The cougar has been characterized as a shy, secretive creature, and many outdoorsmen have spent a lifetime in the western mountains of North America without ever seeing a cougar that was not treed by dogs. In the Rocky Mountains of Idaho, Seidensticker et al. (1973) seldom saw an animal while he was radio-tracking; even at 180 m cougars sneaked away or froze until the investigators passed. Cougars did not avoid the sign of man, but made short-term shifts in their home ranges, depending on the intensity of human disturbance. Conversely, the literature on cougars is full of accounts describing cougar sightings; indeed, most of what is known about this animal comes from such chance encounters.

Several explanations are possible for the large numbers of Manitoba observations. For decades cougars on the western open range and in the mountains were hunted under the bounty system; only recently have they been afforded relief through management as a big-game ani- mal. Even today, only Florida, New Hampshire, New Brunswick, and Manitoba offer this species complete protection. It is possible that individ- uals from the western regions may be more secretive than their counterparts in areas like Manitoba where they are seldom pursued by man or dogs.

Thousands of kilometres of backroads (many new) with limited traffic in wilderness and mixed

1977

lands of Manitoba offer cougars easy travel routes and excellent hunting grounds for the deer that are also attracted to roadsides (e.g., six reports on file were of cougars in active pursuit of deer along roads). More people travelling on an ever-expanding network of backroads in- creases the probability of such a rare event as sighting a cougar. In agricultural regions, cougars crossing flat open areas from one woodlot to another may be spotted from a long distance (particularly transients traversing unknown lands), unlike in rolling or mountain- ous country where visibility is more restricted.

During the 1970s, over 150 cougar sightings were reported in Manitoba. This should not be interpreted as representing 150 different cougars, since it appears that in many cases the same individual or family was observed a number of times. For example, within 40 km south of Swan River, cougars were seen by different people on seven occasions from the summer of 1971 to the fall of 1972, two instances involving an adult with two cubs.

Though the availability of specific prey species has changed in the last 60 years (mule deer and elk to white-tailed deer), there has been no period, perhaps with the exception of the last few years, when there existed a shortage of impor- tant wild prey, not to mention the continual easy access to domestic livestock. Abundant food and cover might imply relatively stable numbers of cougars per unit area in southern Manitoba during this century. The influx of white-tailed deer, however, throughout the former range of mule deer and elk, and far into the boreal forest, apparently has allowed the cougar to expand its distribution to perhaps more than twice its pre- 1940 range. In the last several decades the cougar population in Manitoba has probably been larger than at any other time, and it may still be expanding at some localities. This hypothesis agrees with White’s (1967) view that the turning point in the status of the cougar in Saskatchewan was in the late 1940s, when it changed from being an extremely rare animal and began expanding its range. Our estimate of recent (1970s) numbers of cougars in Manitoba is about 50, based on the wide distribution of reports of single animals and family units. The presence of females with young confirms the existence of a resident population, since only after establishment of a

NERO and WRIGLEY: THE COUGAR IN MANITOBA 39

home range does the cougar end its transient phase and enter the reproductive phase of its life (Seidensticker et al. 1973).

The decline of the deer herds in Manitoba, particularly in the 1970s, to perhaps only 20% of their peak abundance in the 1950s, must have affected cougar numbers and movements. The great increase in cougar sightings in recent years may partly reflect this carnivore’s wider forays in search of its favorite prey, and its attraction to livestock. The continued existence of cougars in Manitoba, as elsewhere, is intricately bound to the proper management of white-tailed deer and the maintenance of suitable habitat.

The cougar was given protection in Manitoba as a rare animal through a regulation to the Wildlife Act passed on 7 January 1974. The Regulation states: “Except under authority of a permit issued by the Minister, no person shall hunt, kill, take or capture a cougar.” Moreover, a person accidentally trapping a cougar must report to provincial wildlife authorities and surrender the animal since, under The Wildlife Act, the carcass is Crown property. The cougar may, however, still be taken by a farmer or stockman to protect his animals if it can be shown that the animal has actually endangered livestock.

Reports from certain western and south- western states show an upsurge in known populations of cougars, e.g., they have doubled in California in the past decade (Williamson 1973), and there is also a renewed interest by resource agencies in determining the status and ecology of this species. In Manitoba, limited steps are now being taken to attempt the capture and radio-tagging of cougars. The results of the present study will hopefully provide a further stimulus to individuals and wildlife agencies in other regions, especially where the cougar is not well known, to investigate and record observa- tions of what Young and Goldman (1946) called, this “mysterious American cat.”

Acknowledgments

Almost all that is known about the cougar in Manitoba is based on observations made through accidental encounters by a large num- ber of persons. We are indebted to these people, and others who directed our attention to them, for providing the basis of this report.

40 THE CANADIAN FIELD-NATURALIST

Members of the Parks Branch (Manitoba Department of Tourism, Recreation and Cul- tural Affairs) and especially staff of the Depart- ment of Renewable Resources and Transporta- tion Services, forwarded reports and names of observers. We are particularly indebted to Regional Biologists L.J. Bidlake, V. F. J. Crichton, D. A. Davies, the late C. C. Holm- strom, R. J. Robertson, and R. C. Thompson. Others who provided special assistance include E. F. Bossenmaier, K. J. Chambers, E. Coulson, A. L. Gardner, H. D. Goulden, C. R. Grondahl, DeReMewHatche MiG. VHornocker Ji le Howard, W. M. H. Koonz, F. L. Lahrman, J. Larcombe, W. H. Longley, D. E. Perry, and M. Shoesmith.

Finally we thank the Manitoba Naturalists Society for helping ensure that the first authenti- cated cougar specimen from the province was deposited in the Manitoba Museum of Man and Nature.

Literature Cited

Abra, M. W. 1974. A view of the Birdtail. History Com- mittee of the Municipality of Birtle. 450 pp.

Bach, R. N. 1943. The mountain lion (Fe/is hippolestes) in North Dakota. North Dakota Outdoors, August. p. 14.

Barnes, C. T. 1960. The cougar or mountain lion. Ralton Company, Salt Lake City, Utah. 175 pp.

Beck, W.H. 1958. A guide to Saskatchewan mammals. Saskatchewan Natural History Society, Special Publi- cation Number |. 52 pp.

Bowles, J.B. 1975. Distribution and biogeography of mammals of lowa. Museum, Texas Tech University, Special Publication 9: 1-184.

Brazier, F. 1960. A mountain lion near Regina. Blue Jay 18: 182-183.

Bue, G. T. and M. H. Stenlund. 1952. Are there mountain lions in Minnesota? Minnesota Conservation Volunteer, September—October 32-37.

Cahalane, V. H. 1961. Mammals of North America. Mac- Millan Company, New York. 682 pp.

Clarke, C. H. D. 1942. Cougarin Saskatchewan. Canadian Field-Naturalist 56(3): 45.

Coues, E. 1897. The manuscript journals of Alexander Henry and of David Thompson. Volume |, The Red River of the North. Ross and Haines Inc., Minneapolis. 446 pp.

Cowan, I. McT. and C. J. Guiguet. 1956. The mammals of British Columbia. British Columbia Provincial Museum, Handbook 11. 414 pp.

Criddle, S. 1929. An annotated list of the mammals of Aweme, Manitoba. Canadian Field-Naturalist 43: 155-159.

Dear, L.S. 1955. Cougar or mountain lion reported in northwestern Ontario. Canadian Field-Naturalist 69: 26.

Vol. 91

Doutt, J. K. 1969. Mountain lions in Pennsylvania? Ameri- can Midland Naturalist 82: 281-284.

Gregg, J. 1974. Cougar collars. British Columbia Western Fish and Wildlife, May. pp. 8-10, 40-44.

Hall, E. R.and K. R. Kelson. 1959. The mammals of North America. Ronald Press Company, New York. 1083 pp. Hubbard, J. 1975. Felis concolor, our big and mysterious cat. New Mexico Wildlife, January-February. pp. 2-6. Jones, J.K., Jr. 1964. Distribution and taxonomy of mammals of Nebraska. University of Kansas Publication,

Museum of Natural History 16: 1-356.

Magnus, L. T. 1956. Mountain lion observation in Lake of the Woods country. Flicker 28(1): 43-44.

Norris-Elye, L.T.S. 1951. The cougar Canadian Field-Naturalist 65: 119.

Riome, S.D. 1973. Evidence of cougars near Nipawin, Saskatchewan. Blue Jay 31: 100-102.

Seidensticker, J. C., 1V., M. G. Hornocker, W. V. Wiles, and J. P. Messick. 1973. Mountain lion social organiza- tion in the Idaho Primitive Area. Wildlife Monographs 35. 60 pp.

Seton, E. T. 1909. Life histories of northern animals—An account of the mammals of Manitoba. 2 volumes. Charles Scribner’s Sons, New York. 1267 pp.

Seton, E. T. 1925. Lives of game animals. Volume | (1). Charles T. Branford Company. pp. 35-136.

Soper, J. D. 1961. Field data on the mammals of southern Saskatchewan. Canadian Field-Naturalist 75: 23-41.

Sutton, R. W. 1960. Seen any cougars around? Wildlife Notebook, Winnipeg Free Press, October 8.

Thompson, E. E. [Seten]. 1886. A list of the mammals of Manitoba. Transactions of the Manitoba Scientific and Historical Society 23: 1-26.

Turner, R. W. 1974. Mammals of the Black Hills of South Dakota and Wyoming. University of Kansas, Museum of Natural History Miscellaneous Publication 60: 1-178.

White, T. 1963. Cougars in Saskatchewan. Blue Jay 21: 32-34.

White, T. 1967. History of the cougar in Saskatchewan. Blue Jay 25: 84-89.

White, T. 1973. Cougar kittens reported near Antler, Saskatchewan. Blue Jay 31: 42-43.

Williamson, L. L. (Editor). 1973. Lion population booms in California. Outdoor News Bulletin, February 16. p. 3.

Wright, B.S. 1959. The ghost of North America—The story of the eastern panther. Vantage Press, New York. 140 pp.

Wright, B.S. 1972. The eastern panther—a question of survival. Clarke, Irwin and Company Ltd., Toronto. 180 pp.

Wright, B.S. 1973. Cougar. Canadian Wildlife Service, “Hinterland Who’s Who.” Information Canada, Ottawa. 4 pp.

Young, S.P. and E.A. Goldman. 1946. The puma— mysterious American cat. The American Wildlife In- stitute, Washington. 358 pp.

in Manitoba.

Received 30 January 1976 Accepted 19 July 1976

Changes in Small Mammal Populations after Clearcutting of Northern Ontario Black Spruce Forest

ARTHUR M. MARTELL! and ANDREW RADVANY/?

'Canadian Wildlife Service, Environment Canada, c/o Great Lakes Forest Research Centre, P.O. Box 490, Sault Ste. Marie, Ontario P6A 5M7

2Canadian Wildlife Service, Environment Canada, 10025 Jasper Avenue, Edmonton, Alberta T5J 1S6

Martell, A. M. and A. Radvanyi. 1977. Changes in small mammal populations after clearcutting of northern Ontario black spruce forest. Canadian Field-Naturalist 91: 41-46.

Abstract. Changes in small mammal populations on upland black spruce (Picea mariana) clearcuts near Manitouwadge, Ontario, were monitored between 1973 and 1975 by live-trapping. Additional clearcuts and uncut stands were sampled by dead-trapping in September 1975. The September density of mice and voles was similar during the first, second, and third year after clearcutting (13.0/ha, 11.6/ha, and 10.0/ha, respectively). Red-backed voles (Clethrionomys gapperi), likely commonin uncut stands but increased on clearcuts until they predominated in the small mammal community by the end of the second harvest, and then declined markedly to rare status. Conversely, deer mice (Peromyscus maniculatus) likely were scarce in incut stands but increased on clearcuts until they predominated in the small mammal community by the end of the second summer after cutting. For less common species clearcut and uncut stands were compared. Meadow voles (Microtus pennsylvanicus) were more common on clearcuts than in uncut stands while the converse was true for rock voles (Microtus chrotorrhinus). Bog lemmings (Synaptomys cooperi) were taken only in uncut stands while heather voles (Phenacomys intermedius) and least chipmunks (Eutamias minimus) were captured only on clearcuts. Clearcutting of upland black spruce forest produced a dramatic change in the composition of the small mammal community but may have caused little change in

density.

Despite the facts that millions of acres of timber are clearcut annually in North America and that small mammals often play an important role in the successful regeneration of forests on clearcuts (see Pank 1974), relatively few studies have been conducted on the effects of timber harvest on small mammal communities. Post- cutting changes have been followed in hardwood forests in New York (Krull 1970) and New Hampshire (Lovejoy 1975), jack pine (Pinus banksiana) forests in Minnesota (Ahlgren 1966) and Manitoba (Sims and Buckner 1973), and Douglas fir (Pseudotsuga menziesii) forests in California (Tevis 1956) and Oregon (Gashwiler 1959, 1970b; Hooven 1969), but the boreal forest, the backbone of the Canadian paper industry, has been largely neglected.

Although there has been a general lack of success with direct seeding as a means of re- generating forests in Canada, the technique offers great potential (see Cayford 1974). The Canadian Forestry Service began investigations of aspects of direct seeding on upland black spruce forests in northern Ontario in 1971 (Fraser 1975; Winston 1975) and requested that the Canadian Wildlife Service carry out co-

41

operative studies of small mammals on the same sites. This is the initial publication of the small mammal investigations.

The study area was located (49°18’N, 85°29’ W) in the Central Plateau Region (B.8) of the Boreal Forest (Rowe 1972) about 32 km northeast of Manitouwadge, Ontario, on the Ontario Paper Company lease. The area is underlain by granite bedrock with pockets of sand and gravel, and soils are generally very thin. Before harvest, upland sites in the area sup- ported mature stands of black spruce (Picea mariana) with a small component of jack pine, paper-birch (Betula papyrifera), and aspen (Populus tremuloides). Dry knolls supported mixed stands of aspen, paper-birch, white spruce (Picea glauca), and balsam fir (A bies balsamea), while wet lowland sites supported stands of black spruce with a small component of white cedar (Thuja occidentalis). The conifers were selectively harvested from the mixed stands leaving aspen, paper-birch, and a thick shrub layer. Some of the lowland sites remained uncut, although others were clearcut except for the cedar. One portion of the study area (Live-Trap Plot 1) (Figure 1) was clearcut in the fall of 1972,

42 THE CANADIAN FIELD-NATURALIST

O 200 meters _—J

Vol. 91

FIGURE |. Map of Manitouwadge, Ontario, study area showing location of live-trap plots (1, 2, 3, 4) and dead-trap lines (A, B, C, D). Note location of clearcuts, uncut black spruce stands (bS), and selective cuts with residual aspen (tA).

and the remainder of the clearcuts and selective cuts in the study area were harvested in the spring and early summer of 1973. All clearcuts were cut by tree-length harvesting methods. Some were scarified using small flanged barrels and spiked anchor chains and subsequently seeded to black spruce.

Methods

Four live-trap plots were established on clearcuts on upland black spruce sites (Figure 1). Plots | (2.39 ha, 161 traps) and 2 (1.49 ha, 100 traps) were established in late August 1973, and Plots 3 (2.91 ha, 196 traps) and 4 (1.43 ha, 96 traps) were established in late August 1974. All plots were established immediately after scari- fication. Trapping points on the plots were 12.2 m apart both within and between rows. A single live-trap was placed within 1.5 m of each trapping point and was covered with a rec- tangular piece of plywood or dry Sphagnum moss and litter. Sherman live-traps were used in

1973 and both Sherman and Longworth live- traps in 1974 and 1975. Traps were baited witha mixture of ground beef suet, raisins, walnuts, peanut butter, oatmeal, and oil of aniseed. A thin slice of apple for moisture and a small handful of terylene fiberfill for bedding were placed in each trap. All traps were checked morning and evening for 10 days, and all newly captured animals except shrews were tagged in each ear with numbered fingerling tags. Shrews were marked with a numbered leg band in 1973 and were color-coded with felt-tipped marking pens in subsequent years. The plots were trapped during the following periods: 1973—Plots | and 2, 28 August-7 September; 1974—Plot 1, 2-12 July; Plots | and 2, 11-21 September; Plots 3 and 4, 28 August-7 September; 1975—Plots | and 2, 3-13 July, 3-13 September; Plots 3 and 4, 15-25 July, 23 August—2 September.

Four dead-trap lines, each consisting of pairs of Museum Special traps set at approximately 15-m intervals and baited with the same bait

1977

mixture as the live-traps, were run 9-12 September 1975 (Figure 1). Each line was run for 72 h and was checked once every 24 h. Line A (46 traps) was set on a dry upland knoll from which the softwood component had been selec- tively cut leaving only aspen, paper-birch, some young white spruce and balsam fir, and a thick _ shrub layer. Line B (54 traps) was set in a black spruce clearcut which was scarified at the same time as Plots 3 and 4 but was on a slightly wetter site. Line C (50 traps) was set in mature uncut black spruce forest, and ran from an upland site down to a lowland site. Line D (50 traps) was set in an upland black spruce clearcut which was not scarified.

Owing to the difficulties in determining densities of small mammals (Krebs and Myers 1974), only a simple estimate of density was calculated from the live-trap data; the total number of individuals captured during a 10-day trapping period was divided by the area of the plot. Because of animal movement into the grid the density is considered to be a maximum figure.

The field work was carried out by several individuals: September 1973—A. Radvanyi, J. Shoup; July 1974—E. Achtemichuk, G. Tessier; September 1974—E. Achtemichuk, J. Shoup; July 1975—D. Fillman, G. Tessier; September 1975—D. Fillman, A. Martell.

Results

Small mammal numbers on all four live-trap plots followed a similar pattern (Table 1), and

MARTELL and RADVANYI: SMALL MAMMALS, NORTHERN ONTARIO 43

therefore the plots will be grouped for dis- cussion. Red-backed voles (Clethrionomys gapperi) predominated in the small mammal community at the end of the first summer after clearcutting (12.3/ha) and in the early part of the second summer (9.6/ha) but declined by the end of the second summer (1.6/ha) and were rare in the third summer (0.5/ha in September). Deer mice (Peromyscus maniculatus), on the other hand, were scarce at the end of the first summer after clearcutting (0.6/ha) but increased markedly by the end of the second summer (12.7/ha on Plots | and 2, and 4.7/ha on Plots 3 and 4) and remained high in the third summer (7.9/ha in September). Meadow voles (Microtus pennsylvanicus), and possibly other voles, and least chipmunks (Eutamias minimus) appeared on the plots the second summer after clear- cutting and their numbers remained relatively stable through the third summer. Masked shrews (Sorex cinereus) were relatively un- common on the plots at the end of the first and third summers after harvest (0.2/ ha and 0.5/ha, respectively) but were abundant at the end of the second summer (5.5/ha). Excluding shrews and chipmunks, the total density of small mammals was similar at the end of the first, second, and third summers after clearcutting (13.0/ha, 11.6/ha, and 10.0/ha, respectively).

The dead-trapping disclosed marked dif- ferences between areas in the composition of small mammal communities (Table 2). The composition of the catch on the scarified clear- cut was similar to that on the live-trap plots

TABLE |—Number per hectare of small mammals on live-trap plots near Manitouwadge, Ontario

Date and plot number

1973 1974 1975 Species September July September July September 1 2 ! 1 2 3 4 1 2 3 4 1 2 3 4 Peromyscus maniculatus — 1.3 3.4 11.3 14.1 5.2 2 4.2 1.3 3.1 1.4 6.7 10.1 65 84 Clethrionomys gapperi 10.5 14.1 9.6 25)| 1.3 1.0 2.1 0.7 0.3 — Phenacomys intermedius! 0.4 0.7 O'S 047) Microtus pennsylvanicus! —_ — ES) 0.8 1.3 28 — 0.8 2s OT 3.5 = 1.3 34 — Microtus chrotorrhinus! 04° — — -- Total mice and voles 10.5 15.4 15.5 14.2 16.7 OHO. Os) SO 4 ON S849 7.5 12.8 10.5 9.1 Sorex cinereus? 0.4 _— 0.4 6.3 5.4 7.6 2.8 — 0.3 a 1.3 _ 0.7 — Sorex fumeus} — _ 7 0.7 Microsorex hoyie 0.8 = — —_ — Blarina brevicauda 0.4 = -- — 0.4 0.7 = — —_ Eutamias minimus — — 0.8 0.7 0.3 — 0.4 = — = 0.4 1.3 0.3 — Grand total 10.9 15.4 15.9 DIPS 22.8. 16.9 9.1 Si8ian 14: 0praell 5.6 10.4 14.8 INES ise al

'Prior to September 1975 M. chrotorrhinus and P. intermedius may have been misidentified as M. pennsyIvanicus

2Prior to September 1975 Microsorex may have been misidentified as Sorex cinereus.

Tentative identification from live specimen; prior to July 1975 S. fumeus and S. arcricus may have been misidentified as S. cinereus

44 THE CANADIAN FIELD-NATURALIST

Vol. 91

TABLE 2—Small mammals captured (number per 100 trap-nights) on dead-trap lines near Manitouwadge, Ontario, 9-12 September 1975

A Selective cut Species (138TN*) Peromyscus maniculatus Sal Clethrionomys gapperi 2.9 Phenacomys intermedius — Synaptomys cooperi = Microtus pennsylvanicus — Microtus chrotorrhinus 28.3 Sorex cinereus = Sorex arcticus _ Microsorex hoyi 0.7 Eutamius minimus Total 37.0

Area (CC B D Mature Scarified Unscarified forest clearcut clearcut (1L50TN) (162TN) (1S50TN) 0.7 5.6 0.7 47 = = zee aes 1.3 1.3 _ — = 8.0 6.0 6.7 — — = 0.6 — = 0.6 aed = = 1.3 13.3 14.8 9.3

*Number of trap-nights (TN) run in each area.

except that meadow voles were more abundant, as would be expected on a wetter, grassier site. The composition of the catch on the unscarified clearcut, however, was conspicuously different from that on the other clearcuts in that deer mice were rare.

The overall abundance of small mammals in mature black spruce forest and on clearcuts was much less than that found in the selective cut in the mixed wood stand. Deer mice were scarce in mature black spruce forest but were common in the selective cut, while red-backed voles were more common in the mature forest. Meadow voles were absent in mature forest and the selective cut but were common on clearcuts, while rock voles (Microtus chrotorrhinus) were abundant in the selective cut and common in the mature black spruce forest but were rare on clearcuts (Tables | and 2). Bog lemmings (Synaptomys cooperi) were restricted in their distribution to uncut black spruce forest while heather voles (Phenacomys intermedius) and least chipmunks were captured only onclearcuts (Tables 1 and 2). Shrews may have been more common on clearcuts than in uncut stands but this is inconclusive because of the trapping methods used. Pitfall traps would probably have given better estimates of shrew abundance.

Discussion The most noticeable change in the small

mammal community after clearcutting of up- land black spruce forest was in the composition rather than in the density. Red-backed voles likely were common in uncut stands, predom- inated on clearcuts through the early part of the second summer after harvest, and then declined rapidly until they were rare by the end of that summer. Conversely, deer mice likely were scarce in uncut stands, increased during the second summer after clearcutting, and predom- inated in the small mammal community by the end of that summer. A similar pattern of decrease in red-backed voles and increase in deer mice after clearcutting has been observed in Douglas-fir forests (Gashwiler 1959, 1970b; Hooven 1969; Tevis 1956) and in jack pine forests (Sims and Buckner 1973); the pattern was also present after fire in black spruce — jack pine — paper-birch forests in Minnesota (Kreft- ing and Ahlgren 1974). Other studies have found that red-backed voles may remain rare or absent on clearcuts for 4 to 10 years after harvest (Gashwiler 1967, 1970a, b; Hooven 1969; Kreft- ing and Ahlgren 1974) but become common sooner if thick groundcover becomes established (Ahlgren 1966; Lovejoy 1975). Red-backed vole microdistribution has been reported to show a strong correlation with the amount of debris cover and evergreen shrub cover (Lovejoy 1975; Miller and Getz 1972, 1973). It is likely, therefore, that it is the lack of cover that makes

1977

clearcuts unsuitable for red-backed voles. Deer mice, on the other hand, are able to exploit the relatively barren habitat of clearcuts and either increase from the small stock remaining after clearcutting or to invade the clearcuts from surrounding ‘habitats where they are more abundant, such as the selective cuts adjacent to the upland black spruce clearcuts.

The increase in the deer mouse population was probably a response to clearcutting rather than to the decrease in red-backed voles since no strong behavioral competitive interaction has been shown between those species (Grant 1970) and there is little overlap in food habits (Dyke 1971; Wilhams 1959). The unusually low numbers of deer mice on the unscarified clear- cut, however, is unexplained. It suggests that where an increase in deer mice occurred it was in response to scarification. There is additional support for that suggestion: in September 1974 deer mice were more abundant on Plots | and 2, which had been scarified the year before, than on Plots 3 and 4, which had just been scarified. It is unlikely that the lower densities of deer mice on Plots 3 and 4 were due to the probable adverse effects of recent scarification, as red-backed vole numbers did not show a similar depression. Because Plots | and 2 were seeded in fall 1973 and Plots 3 and 4 in fall 1974, it might be argued that the deer mice were actually responding to seeding. But the seeds used on Plots | and 4 were treated with an apparently effective rodent repellent (Fraser 1975). More investigation is necessary to clarify the situation.

Meadow voles were taken only on clearcuts at least | year old and varied in numbers among the areas sampled. Microdistribution of meadow voles has been reported to correlate with moisture and graminoid vegetation cover (Getz 1961, 1970). It is likely, therefore, that they appear on clearcuts once sufficient cover has been established and in densities proportional to the amount of moist, graminoid cover present. Bog lemmings and heather voles also displayed narrow tolerances and were found only in their preferred habitats, moist coniferous forest and dry shrubby areas, respectively (Foster 1961; Getz 1961). The high density of shrews found on the clearcuts in September 1974 may indicate a general population peak such as has been described in other northern areas (Buckner 1966;

MARTELL and RADVANYI: SMALL MAMMALS, NORTHERN ONTARIO 45

de Vos 1957) and is probably not a response to clearcutting. Shrew densities may have been higher on clearcuts than in uncut stands owing to the probable increase in invertebrates on clear- cuts (cf. Lovejoy 1975). The great numbers of rock voles found in the selective cut and the mature black spruce stand in September 1975 is interesting. Because no trapping had been done in uncut stands in the study area previously, it is not known whether the observed abundance was ‘normal’ or a cyclic population ‘peak,’ if in fact rock voles do cycle. Also the population was found in an area where their supposed preferred or exclusive habitat, rocky outcrops and talus slopes (Burt 1957; Peterson 1966), was absent.

Clearcutting of upland black spruce forest in northern Ontario altered the environment so that clearcut sites were less desirable than uncut sites for red-backed voles, rock voles, and bog lemmings, although the opposite was true for deer mice, meadow voles, heather voles, and least chipmunks. Shrews may also have been more common on clearcuts. There appeared to be little change in overall small mammal density due to clearcutting.

The marked increase in deer mice after clearcutting is of particular interest to foresters because that species is considered a much more serious seed predator than the red-backed vole (see Pank 1974). Although there is no proof that either species is a serious predator of black spruce seeds, the initial data are suggestive (Fraser 1975). Assuming that black spruce seed predation occurs, the data presented in this paper suggest that on upland black spruce sites in northern Ontario seeding should be done as soon after clearcutting and site preparation as possible in order to minimize the chances of seed losses to deer mice. Also, spring seeding should be more successful than fall seeding, a suggestion that is substantiated by the initial studies by Fraser (1975) and by studies in other areas (Radvanyi 1970, 1971).

Acknowledgments

It was largely through the efforts of J. W. Fraser that the Canadian Wildlife Service became involved in forest regeneration studies on upland black spruce sites in northern Ontario; for that, and for his continued coopera- tion and assistance during the study, we are

46 THE CANADIAN FIELD-NATURALIST

grateful. We are indebted to J. H. Cayford, Director of the Great Lakes Forest Research Centre, and to the Ontario Ministry of Natural Resources and the Ontario Paper Company for their cooperation. We thank E. Achtemichuk, D. Fillman, J. Shoup, and G. Tessier for their assistance in the field, and J. Shoup for preparing the map of the study area. This paper benefited from critical review by D. R. Flook and D. A. Welsh, Canadian Wildlife Service, Ontario Region.

Literature Cited

Ahlgren, C. E. 1966. Small mammals and reforestation following prescribed burning. Journal of Forestry 64: 614-618.

Buckner, C. H. 1966. Populations and ecological relation- ships of shrews in tamarack bogs in southeastern Manitoba. Journal of Mammalogy 47: 181-194.

Burt, W. H. 1957. Mammals of the Great Lakes region. University of Michigan Press, Ann Arbor, Michigan. 246 pp.

Cayford, J. H. (Editor). 1974. Direct seeding symposium, Timmins, Ontario, September 11, 12, 13, 1973. Canada Department of the Environment, Forestry Service Publi- cation Number 1339. 178 pp.

deVos, A. 1957. Peak populations of the masked shrew in northern Ontario. Journal of Mammalogy 38: 256-258.

Dyke, G. R. 1971. Food and cover of fluctuating popula- tions of northern cricetids. Ph.D. thesis, University of Alberta, Edmonton, Alberta. 245 pp.

Foster, J. B. 1961. Life history of the phenacomys vole. Journal of Mammalogy 42: 181-198.

Fraser, J. W. 1975. Direct seeding black spruce—is it feasible? Jn Black spruce symposium, Thunder Bay, Ontario, September 23, 24, 25, 1975. Canada Department of the Environment, Forestry Service, Great Lakes Forest Research Centre Symposium Proceedings O-P-4. Pp. 140-155.

Gashwiler, J. S. 1959. Small mammal study in west-central Oregon. Journal of Mammalogy 42: 128-139.

Gashwiler, J.S. 1967. Conifer seed survival in a western Oregon clearcut. Ecology 48: 431-438.

Gashwiler, J.S. 1970a. Further study of conifer seed survival in a western Oregon clearcut. Ecology 51: 849-854.

Gashwiler, J.S. 1970b. Plant and mammal changes in a clearcut in west-central Oregon. Ecology 51: 1018-1026.

Getz, L. L. 1961. Factors influencing the local distribution of Microtus and Synaptomys in southern Michigan. Ecology 42: 110-119.

Getz, L.L. 1970. Influence of vegetation on the local distribution of the meadow vole in southern Wisconsin. University of Connecticut Occasional Papers (Biological Science Series) 1: 213-241.

Grant, P. R. 1970. Experimental studies of competitive interaction in a two-species system. II. The behavior of Microtus, Peromyscus and Clethrionomys species. Animal Behavior 18: 411-426.

Vol. 91

Hooven, E. F.: 1969. The influence of forest succession on populations of small animals in western Oregon. Jn Wild- life and reforestation in the pacific Northwest. Edited by H.C. Black. Proceedings, Symposium, 1968 (Oregon State University, School of Forestry, Corvallis). Pp. 30-34.

Krebs, C. J. and J. H. Myers. 1974. Population cycles in small mammals. Advances in Ecological Research 8: 267-399.

Krefting, L. W. and C. E. Ahlgren. 1974. Small mammals and vegetation changes after fire in a mixed conifer- hardwood forest. Ecology 55: 1391-1398.

Krull, J. N. 1970. Small mammal populations in cut and uncut northern hardwood forests. New York Fish and Game Journal 17: 128-130.

Lovejoy, D.A. 1975. The effect of logging on small mammal populations in New England northern hard- woods. University of Connecticut Occasional Papers (Biological Science Series) 2: 269-291.

Miller, D. H.and L. L. Getz. 1972. Factors influencing the local distribution of the redback vole, Clethrionomys gapperi, in New England. University of Connecticut Occasional Papers (Biological Science Series) 2: 115-138.

Miller, D. H.and L. L. Getz. 1973. Factors influencing the local distribution of the redback vole, Clethrionomys gapperi, in New England. II. Vegetation cover, soil moisture, and debris cover. University of Connecticut Occasional Papers (Biological Science Series) 2: 159-180.

Pank, L. F. 1974. A bibliography on seed-eating mammals and birds that affect forest regeneration. United States Department of the Interior, Fish and Wildlife Service, Special Scientific Report, Wildlife Number 174. 28 pp.

Peterson, R.L. 1966. The mammals of eastern Canada. Oxford University Press, Toronto. 465 pp.

Radvanyi, A. 1970. Small mammals and regeneration of white spruce forests in western Alberta. Ecology 51: 1102-1105.

Radvanyi, A. 1971. Lodgepole Pine seed depredation by small mammals in western Alberta. Forest Science 17: 213-217.

Rowe, J.S. 1972. Forest regions of Canada. Canada Department of the Environment, Forestry Service Pub- lication Number 1300. 172 pp.

Sims, H. P. and C. H. Buckner. 1973. The effect of clear cutting and burning of Pinus banksiana forests on the populations of small mammals in southeastern Manitoba. American Midland Naturalist 90: 228-231.

Tevis, L., Jr. 1956. Responses of small mammal popula- tions to logging of Douglas-fir. Journal of Mammalogy 37: 189-196.

Williams, O. 1959. Food habits of the deer mouse. Journal of Mammalogy 40: 415-419.

Winston, D. A. 1975. Black spruce seeding experiments in the Central Plateau Section B.8, Manitouwadge, Ontario. In Black spruce symposium, Thunder Bay, Ontario, September 23, 24, 25, 1975. Canada Department of the Environment, Forestry Service, Great Lakes Forest Research Centfe Symposium Proceedings O-P-4. Pp. 125-139.

Received 16 July 1976 Accepted 4 October 1976

Morphological Parameters and Spring Activities in a Central Ontario Population of Midland Painted Turtle, Chrysemys picta marginata (Agassiz)

T. H. WHILLANS! and E. J. CROSSMAN23

1Geography Department, University of Guelph, Guelph, Ontario

Present address: Department of Geography, University of Toronto, Toronto, Ontario M5S IAl 2Department of Ichthyology and Herpetology, Royal Ontario Museum, Toronto, Ontario M5S 2C6

3Correspondence should be directed to E. J. Crossman.

Whillans, T. H. and E. J. Crossman. 1977.

Morphological parameters and spring activities in a central Ontario popula-

tion of midland painted turtle, Chrysemys picta marginata (Agassiz). Canadian Field-Naturalist 91(1): 47-57.

Abstract. Chrysemys picta marginata in a north-central sector of its range exhibits values for standard taxonomic character- istics that have wide individual variability. A population at Nogies Creek, Ontario did not exhibit values which might, from the literature, be expected (“typical”) for this subspecies in that part of its range. Individual sizes do approach maxima for midland painted turtles. Nesting activities are described for the population. Photoperiod appears to be an instrumental guideline for timing daily and anriual nesting activities. Other environmental factors may temper its effect. Turtles might

locate the nest sites by a simple form of homing.

Much information has been published within the past 25 years on the midland painted turtle, Chrysemys picta marginata (Ernst 1971c). Des- pite this animal’s popularity as an experimental subject, it has received geographically extensive attention only in areas of intergradation. Few data are available from any northern popula- tions. Information on the subspecies must often be extrapolated from discussions of the species, Chrysemys picta (Schneider).

The present study was undertaken for the following reasons:

(1) to provide some basic morphological and observational data on C. p. marginata in the north-central portion of its range. These sta- tistics can then be compared with findings of authors such as Hartman (1958), Pough and Pough (1968), and Ernst (1970a) in the eastern and southern zones of intergradation.

(2) to describe the spring activities of the midland painted turtle in this north-central study area. Sexton (1959) divided the annual activity cycle of Chrysemys into five seasons. This project focuses on behavior in the late vernal and early aestival periods, with the hope of delineating nesting and associated move- ments.

Study Area Chrysemys p. marginata was collected at the

47

Nogies Creek Fish Sanctuary in Peterborough County, Ontario (Figure 1). Previously des- cribed by Crossman (1956) and Muir (1963), the sanctuary consists of a meandering 6.4-km stretch of drowned stream, the lower portion of which widens into a shallow 32.4-ha lake. Most open water is from 1.5 to 2.0 m deep although several spots drop to 5.0 m. Aquatic vegetation emerges early in June and by the end of that month has reduced the open water to a narrow channel. The surrounding land is predominately a shallow gravelly till over frequently exposed bedrock. Mixed forest lines most of the shore- line.

Nogies Creek Research Station trap-net catch records dated 1951 reveal painted turtle catches approximately similar to those of the present. Personal observations (Crossman) have recog- nized no major changes in turtle abundance through the ensuing 25 years. This suggests a long-standing, stable population of C. p. marg- inata. At least 20 species of fishes, 61 varieties of aquatic plants, and 2 other species of turtle, Blanding’s turtle, Emydoidea blandingi (Hol- brook), and common snapping turtle, Chelydra s. serpentina (Linnaeus), cohabit the sanctuary.

Materials and Methods

Between mid-May and early July of 1974 turtles were captured in trap-nets and hoop-nets

48

THE CANADIAN FIELD-NATURALIST

Vol. 91

[-]) «Cp. marginata ; = Cp. belli : 2 iasceecereces ; : ROA C.p. picta : a 2 => 3 ° QUE, 2 m3 ps) Os 2 7 77 v ° SS 2 p re, = % BS es ay Fe BER ;-—WE — ay ATS Le meee: aa A x wate = —— es fat Ly sug eee —— ZF, SSS vou sek = RATS = — rakes, + = geese se. F =~ LORY z = oan Wa ®, + SA i reeas ( Br ae, - Soff EER wis eeeeeses =. — RES 5 Seaneee : MICH. EEE : EOE 4 _- pee = : ce : (as) Kee SRR Coenen LEER : : : Lee Ogee Sey, : ‘ awe coaee = = : os : : cy : PREETI rene = Ln eon on into pm on tt r : BEE = H 1 eee eee. z ee Og Se La, — 5 — ‘ PALS = : : 4 t : t EEO 4 : : — Bes OH | Q ¥ . os z Dee ae, 7 \ t i omen ge Sas Ss Sm, {9 : Se cos Se l ai —— = Cl CT 4 J MES. = ; rT. oF OWA == —— : = =f a KY. 4 0 50100 200 300 =o oe : ¢ a enn s BP KILOMETERS

FIGURE 1. Ontario study location in the north-central portion of the range of C. p. marginata. Map adapted from Conant

(1975). Intergradation occurs in zones of overlap.

with occasional opportunistic hand sampling. The nets were regularly relocated in a total of 25 sites within the lower sanctuary. Capture by hand was necessary only for terrestrial samples. Carapace total length and carapace curved length were measured with a flexible metal tape rule accurate to 1.0 mm. Weight was recorded by placing the animal on its back in a plastic bucket which was then set on a hanging spring scale precise to 25 g. Sex was determined from the length of the foreclaws. In 1974 each turtle was numbered on the plastron with indelible red ink and when the first animal with an illegible number was recaptured, after approximately 2 months, trapping was discontinued. In the first 3 weeks of June 1975 turtles taken in the nets were marked and released but the main emphasis at that time was observations on shore. In 1975, numbering of individual turtles was accom- plished by notching the marginals with a hacksaw blade (Cagle 1939). Each marginal had been assigned a number so that a notched

marginal, or combination of marginals, served to identify turtles in a more permanent manner than the previous year’s ink markings.

The three characters used by Hartman (1958) to distinguish C. p. marginata from C. p. picta (Schneider) were measured with calipers exact to 0.01 mm. Alignment of central and lateral laminae was measured and calculated as des- cribed by Hartman (1958) and later by Pough and Pough (1968). One hundred percent dis- alignment describes exact alternation of the laminae as is “typical” in C. p. marginata, while 0% represents the linearly catenated seams “typical” in C. p. picta. The plastral figure, if present, was copied directly onto tracing paper to allow for later experimentation with different plastral figure measurements. Eventually figure length, greatest figure width, longest figure extension along a seam, and plastron length were measured to the nearest 0.5 mm. Figure length and greatest figure width were plotted against plastron length, and longest seam-figure

1977 WHILLANS and CROSSMAN: MIDLAND PAINTED TURTLE, CENTRAL ONTARIO 49

TABLE |—Statistics relating to shell markings of Chrysemys picta marginata in Nogies Creek, Ontario, 1975. n, number in sample; x, mean of the “characteristic” expressed as a percentage for the total of the relevant sample population; SD, standard deviation (one); SE, standard error (one); , population mean estimated by X, subscript 1 represents the male sample, sub-

Characteristic Sample composition n x Plastral figure Total sample 47 62.98 length/ plastral Males 222) 64.66 length Females 23 64.13 Greatest plastral Total sample 47 46.85 figure width/ Males 22 43.00 greatest plastral Females 23 50.14 width Longest seam- Total sample 47 20.34 figure-extension/ Males 22 19.10 greatest figure Females 23 22.26 width Light margin/ Total sample 50 1.04 total length Males 23 0.93 Females 26 1.13

'Critical value 95% confidence = 1.960.

extension against greatest figure width (see Table 1). Those ratios test the length relation- ships that interested Hartman (1958), the figure width characteristics that concerned Bishop and Schmidt (1931), and the figure extension traits that Carr (1952) found to be important. Hart- man (1958) and Ernst (1970a) indicated a sub- specific differentiation in the width of the anterior light margin on the third lateral. This was measured. The total length (anterior to posterior) of each of the carapace and plastron was recorded. Shell abnormalities were also noted.

Observations of nesting activities were con- centrated at four well used sites. A gravel road runs parallel to the west shore of Nogies Creek for a distance of 3.2 km. The turtles had to cross the road to reach three of the nesting locations and were easily seen. Observations started at 0700 hours and ended at dark since there were no references to night activity by C. p. marginatain the literature surveyed. In the beginning the road was traversed every half hour and observations were made on all turtles that could be located from the road. As a result it became obvious that there were daily peak periods of activity and

script 2 represents the female sample

SD SE 95% confidence Test hypothesis! Mi-uw2 = 0

16.30 2.38 58.19-67.76

8.32 1.77 0.0555 Accept 45.28 9.44

9.16 1.34 44.16-49.54

9.70 2.01 2.7401 Reject 7.58 1.58

8.67 1.26 17.79-22.88

6.66 1.42 1.2951 Accept 9.52 1.98

0.36 0.05 0.94-1.14

0.38 0.08 1.8554 Accept 0.35 0.07

periods of lesser activity. Subsequently activity was monitored every half hour during the known peak periods and at hourly intervals otherwise. The duration of each observation (on one or more than one turtle) varied but the road was not left unmonitored for longer than the intervals given above. The fourth nesting location, having poor accessibility, was visited irregularly.

Date, time of day, location, identifying number, and morphological characteristics were recorded for each turtle. When an animal was found to be nesting, its activities were observed from a place of hiding or with binoculars from a distance. Measurements were made after nest- ing, or when the turtle was returning to the water.

Nests were measured (diameter of neck, depth of main chamber, depth to bottom of main chamber). The substrate was described, distance from shore estimated, and terrain noted. If a turtle was seen at a nest which was later covered over, the nest was unearthed. From any nest with eggs, clutch size, depth to eggs, and egg size (length and width) were recorded.

Air temperature was measured daily with a maximum-minimum thermometer. Daily max-

50 THE CANADIAN FIELD-NATURALIST

imum and minimum air temperatures from the nearest official Climatological Station in Cobo- conk were obtained through the Meteorological Applications Branch of Environment Canada. The Corn Heat Units (CHU) were calculated from these (see Brown 1972). Water temperature was not considered because of an incomplete record at the sanctuary; however, the close correlation between air and water temperatures should minimize the importance of this omission (see McCombie 1959).

Measurements and Observations Standard Measurements

A sample of 50 turtles from Nogies Creek was tested for disalignment of the lateral and central laminae seams. The mean disalignment was 86.4% with a standard deviation of 11.3 and standard error of approximately 1.6. This is in accordance with the findings of Pough and Pough (1968) and Hartman (1958), and is well above the minimum 55% disalignment set by Ernst and Ernst (1971) for the midland painted turtle. Figure 2 shows that of populations studied to date, the Nogies Creek population is

“—

CAPE COD, MASS.

(H 12)

Vol. 91

perhaps the closest to expected findings for C. p. marginata populations. The wide range of 70 disalignment percentage points (36-102) is partly explained by one deviant individual (36%). The next value higher than 36% was 73%.

Plastral figure measurements are summarized in Table 1. The mean value for figure length as a percentage of the plastral length is 62.9. Greatest plastral figure width averages 46.8% of the greatest plastral width. The mean value for longest seam-figure extension as a percentage of the greatest figure width is 20.3%.

Since the Nogies Creek plastral figures were all recorded on tracing paper, they can be compared to the diagrams of C. p. marginata, C. p. bellii (Gray), and an intermediate given by Bishop and Schmidt (1931). A visual inspection of their sketches reveals that in a Nogies Creek sample of 50 turtles the majority (38 or 76%) had what seems to be C. p. marginata plastral figures. None displayed the markings of C. p. bellii, but nine (18%) had intermediate character- istics.

No one has yet been able to devise a totally

PERCENT 50

CUTTYHUNK, MASS, (P 14) SOUTH AMHERST, MASS. (H 19) NANTUCKET, MASS. (Pp 25) LONG ISLAND, NY. (Pp 73) Northern NEW JERSEY (P 26) DRYDEN, farmpond NY, (H 21) FISH HATCHERY, ITHACA NY. (H 20) FLINT, MICHIGAN (H)

SODUS BAY, NY. (H 25)

NOGIES CREEK, ONT, (50)

P—Pough and Pough (1968) °

| H— Hartman (1958)

eo

50 PERCENT

FIGURE 2. Percentage of disalignment of central and lateral seams for several populations depicting range, mean, one standard error and one standard deviation on each side of the mean.

1977

reliable statistical method for expressing the differences between the plastral figures of the two subspecies, because of the high degree of individual variation. Bishop and Schmidt (1931) used the width of the plastral figure expressed as a percentage of the plastral width. According to their study, the mean value for C. p. marginata was 36%, for C. p. bellii 74%, and for inter- mediates 55%. The Nogies Creek turtles, with a value of 46.8% (Table 1) lie about halfway between the value for intermediates and that for C. p. marginata. Only eight (16%) individuals had plastral figures with widths less than 36% of the plastral width, and none were in excess of 74%.

Three of the Nogies Creek turtles had no plastral figure at all. This characteristic is normal in C. p. picta. The mean value of figure length as a percentage of plastral length for Nogies Creek turtles is 62.9% (Table 1). There are no comparable figures in the literature. Bleakney (1958) indicated two intergrade figure traits: (1) dark spots on the midline, and (2) smaller, often lighter figures, symmetrically distributed about the midline and having blurred edges. None of the Nogies Creek sample dis- played dark spots on the midline and only four turtles were faint in color and had figure lengths

WHILLANS and CROSSMAN: MIDLAND PAINTED TURTLE, CENTRAL ONTARIO 51

less than 50% that of the plastron.

The longest seam-figure extension was cal- culated as a percentage of the greatest figure width (20.3) with little utility. This relationship did not support the visual impression of the plastral figure as did the aforementioned com- parisons involving figure length and width; however, it has been presented in Table | as the best measure of seam extensions. Other traits compared less successfully to longest seam- figure extension were plastral length, figure length, and greatest plastral width.

Hartman (1958) plotted the width of the light margin against carapace length. Nogies Creek painted turtles have been similarly character- ized and compared with Hartman’s results in Figure 3. The Ontario turtles align themselves with the New York samples of C. p. marginata and are clearly distinguishable from the Massa- chusetts intergrade population (C. p. marginata XC. p. picta).

All three of the ventral figure relationships, and that between the light margin and carapace length, were tested for sexual dimorphism. The width of the plastral figure, when expressed as a percentage of the width of the plastron, was the only one of these characteristics having two Statistically different sexual values (see Table 1).

a S.Amherst, Mass. (Hartman, 1958) © Dryden, N.Y. (Hartman, 1958) @® Nogies Creek, Ont.

5 — E

4 z O a < = 3t

a = | <= oe me ca ae oO | x a al a ) = | : 00 0

0) 25 50 75

100 125 150 175

Te eee ea aca|

200

CARAPACE LENGTH (mm) FIGURE 3. Width of the light margin of the third lateral plotted against carapace length. Adapted from Hartman (1958).

52 THE CANADIAN FIELD-NATURALIST

These calculations are only slightly significant (test hypothesis value = 2.74, 95% confidence = 1.96) and because of the nonsignificant values for each of the other three shell-marking characteristics, sexual dimorphism was not pursued any further.

Other Measurements and Morphometric Obser- vations

Frequency polygons for length and weight in the sample of Nogies Creek turtles are given in Figure 4. A total of 195 turtles was captured. They ranged in length from 105 to 181 mm and in weight from 175 to 675 g. According to Ernst (1971a), male turtles in Pennsylvania usually mature at a plastral length of about 70 mm, in their fourth year, and females at approximately 100 mm, in their fifth year. If this is applied to the Ontario population it appears that netting was selective for mature painted turtles, espe- cially since there was no problem sexing even the smallest animals. Females were larger than males.

Fifty-one turtles were examined in 1975 for the humped-back condition (“kyphosis”) des- cribed by Ernst (1971b). None displayed the trait. An additional 195 turtles from 1974 were tested for deviation from a straight-line relation-

Vol. 91

ship between the total length of the carapace and the curved length of the carapace. No obvious deviants were noted in these Nogies Creek turtles. Ernst (1971b) noted that 0.5% of his sample at the White Oak Bird Sanctuary in Pennsylvania exhibited kyphosis:

In a subsample of 51 turtles from Nogies Creek, exterior deformities were recorded: 20 (39%) had deformites of the following kinds. Plastral abnormalities were the most numerous. Ten of the 20 animals (19.5% of 51 sample turtles) had misshaped plastrons. Ernst (1971b) found only one occurrence of this peculiarity in his collection of 929 turtles. In the White Oak study, carapacial abnormalities were found on 6.2% of the deformed turtles. This variation comprised 13.7% of the Nogies Creek sample. Sixteen percent of the 5! Ontario turtles examined displayed unusual appendages. Com- binations of the above deformities were not uncommon in either Ernst’s or the present studies.

Spring Movements: Aquatic

Figure 5 summarizes the numbers of C. p. marginata captured per net, per day, in 1974. A comparison of the daily air temperatures for the same period of time, also in Figure 5, reveals a

CARAPACE TOTAL LENGTH (mm)

! female (s) male (s}

150 160 170 180 = ea

NUMBER OF TURTLES ie

al Ke 200 250 300 350 400 450 500 550 600 650 700 WEIGHT (g)

FIGURE 4. Lengths and weights of male and female C. p. marginata from Nogies Creek.

1977

----- Max. temperature —— Min. temperature seoeee No. of turtles = ae Data missing

(G0) {o)

NO (eo)

AIR TEMPERATURE (°C)

22 “BS 30 1 5 10 15

WHILLANS and CROSSMAN: MIDLAND PAINTED TURTLE, CENTRAL ONTARIO 53

NUMBER OF TURTLES PER NET

20 . 25 1 5 10

—M AY —e JUNE JULY

DATE

1974

FIGURE5. Daily maximum and minimum air temperatures plotted with the numbers of turtles caught per net, calculated per

24-h period in 1974.

direct relationship between air temperature and turtle movement reflected in the number cap- tured. But, there often was a time lag between peaks of temperature and activity. There was also a coincidental increase in temperature and movement immediately prior to the initiation of nesting activities. Extensive cloud cover was observed (no data) to have a negative influence on turtle activity.

Turtles were first seen nesting on 6 June in 1974 and 12 June in 1975. At the initiation of nesting in 1974 approximately 515 Corn Heat Units (CHU) had accumulated since mid-April (calculated from records at a climatological station 19.3 air km to the northwest). Nesting in 1975 started when 804 CHU had been recorded.

In the aquatic netting of 1974, 174 different C. p. marginata were captured, measured, and returned to exactly where caught, before plastral numbering became illegible. Of these, 43 (24.7%) were recaptured, 32 (18.3%) only once, seven (4.0%) twice, two (1.1%) thrice, one 8 times, and one 10 times, for a total of 244 captures. Twenty-

six of the 43 recaptured turtles had changed position between captures.

In the total of 244 aquatic captures between 22 May and I1 July 1974 there were 159 males and 85 females; the sex ratio was 1.87:1. Because some turtles were captured more than once, actually only 174 individual C. p. marginata were caught in the nets. This included 104 males and 70 females for a sex ratio of 1.49:1. Single captures were recorded for 131 turtles with a 1.34:1 sex ratio. This contrasts sharply with the sex ratio of 2.07:1 for individuals displaying multiple recaptures. When the number of cap- tures per sex are summed, regardless of the number of turtles involved, the sex ratio be- comes 3.67:1.

Spring Movements: Terrestrial

All 30 terrestrial captures, involving 25 C. p. marginata, were females. Three of the five recap- tures occurred on the day after the initial capture. These three recaptured individuals were all discovered within 5 m of the point of original

54 THE CANADIAN FIELD-NATURALIST

capture, one no more than 0.5 m from the first site. Another of the five recaptures was found after 2 days about 15 m from the initial location. The last was relocated 8 days later approxi- mately 150 m from the original site.

A large number of empty nest holes was found in the most frequented nesting areas. In 1974 an unrecorded number of unmarked turtles was observed digging nests which they later aban- doned as empty holes. Sometimes these con- tained obstacles such as rock; in others there was no apparent physical restriction.

Only three turtles were captured both in the water and on land. One had travelled 250 m from the initial site of netting to an adjacent part of the shore. It was found 100 m from the water by a sandy roadside bank.

Two turtles were discovered on land 850 m and 1075m from the original location of capture. These represent the farthest distances travelled by any of the marked animals. The two were found within 30 m of each other at one of the most-used nesting sand banks. One had been previously captured in the water 1075 m to the north along an irregular shoreline. The other was later captured in a net 850 m to the south along a similar shore. From the creek there was no apparent visual clue to the location of the sand bank owing to the high bushes which line most of the water’s edge.

Nesting Activities

Nesting behavior in painted turtles has pre- viously been described: C. p. marginata (Hart- weg 1944; Carr 1952), C. picta (Babcock 1919), C. p. bellii (Legler 1954; Mahmoud 1968). Al- though direct observations at Nogies Creek largely confirmed these other findings, three additional points were recorded. Four turtles were observed from the time of initial search for a suitable nest site to the final covering over of the eggs. Nesting behavior for an additional dozen animals was noted in part.

(1) A female C. p. marginata would move out of the water at a steady pace and with its head level. It would stop upon reaching denser vegetation, lift its head and remain static, or proceed erratically until a new direction was established. This behavior continued until the turtle came to a potential nesting area. Then it would lower its head until almost touching the ground and creep

Vol. 91

ahead slowly as if sensing something. Occasion- ally it would scrape the ground with its claws. The site eventually was accepted or rejected. (2) The periodic discharge of liquid from the anus during digging was described by Legler (1954). At Nogies Creek turtles measured prior to laying eggs discharged liquid, but any cap- tured upon completion of a confirmed nesting had no liquid discharge. Turtles continued nesting activities even after discharging their liquid. Thus the liquid, which seemed to aid in digging, is probably not a necessity. (3) When laying was completed, the female turtle would cover the nest hole using its hind legs and tamp the surface down with its plastron. One turtle was observed dragging some nearby (one shell length) loose leaves over the nest with its front legs. Nests in grass were almost in- distinguishable even immediately after their formation. A rain usually obliterated all traces. Seventeen female C. p. marginata were found moving on land and away from the water. They were all discovered between 1500 and 1915 hours (1640 hours mean). The duration of their activities varied from | to 4h with the largest recorded capture being at 2100 hours. Turtles were located anywhere from | to 200 m away from the water on shore (mean 46 m). Approx- imately 37 nests, however, ranged between 2 and 50 m from the shoreline, averaging 20 m.

Clutch size in five nests varied from 6 to 9 eggs, with a mean of 7.2. This is within the 3 to 11 egg range listed by Carr (1952). Immediately after laying, the eggs had an average width of 18.5 mm (range 16.9 to 19.5 mm, n= 14) and a mean length of 30.8 mm (range 26.7 to 50.0 mm, n = 14). One nest examined after it was incubated for 73 days contained five apparently healthy eggs (plus four mutilated) which averaged wider but shorter than those newly laid (width: mean 20.5 mm, range 18.1 to 22.2; length: mean 29.0 mm, range 25.5 to 31.1). The comparable ranges listed by Carr (1952) are length 28.6-31.8 mm, and width 17.5—20.6 mm.

Mean nest dimensions were as follows: dia- meter of neck, 29.1 mm (n = 3); diameter of main chamber, 65.2mm (n=7); depth to eggs, 25.0mm (n=1); depth to bottom of main chamber, 70.6 mm (n= 6). Ernst and Barbour (1972) described nests of C. picta as having the

1977

following dimensions: main chamber 65-72 mm, neck diameter 41-51 mm, and depth 99-111 mm. Most nests were situated in sandy loam, gravel, or sandy gravel. The ground was clear, or hada grass cover of up to 30 cm in height. All received the sun most of the day and there were few not on level ground or on the lower slopes of an east- facing bank. From the data of this study it is impossible to estimate nesting success. Pred- ators were numerous, racoons, people, and larval insects being common.

No vernal emigration as reported by Sexton (1959) was noted in Nogies Creek. Constant water levels regulated by a dam at the lower end of the sanctuary may account for this.

Discussion Standard Measurements

Bleakney (1958) hypothesized that the north- ward postglacial dispersal of C. p. bellii and C. p. dorsalis (Agassiz) resulted in an intra- specific hybridization around St. Louis, Mis- souri. The subspecies C. p. marginata emerged and subsequently extended its range to its present status.

The Nogies Creek turtles clearly display cara- pacial disalignment and light margin width “typical” of C.p. marginata. The plastral figures, however, seem to be wider and perhaps longer than might be expected. This does not necessarily indicate intergrade characteristics of the north-central subspecies with another. It could be evidence of the C. p. bellii influence in Bleakney’s postulated C. p. bellii X C. p. dor- salis origin of C. p. marginata. Similarly, al- though the lack of plastral figure or its reduced size is a trait documented in intergrade C. p. marginata X C. p. picta, its occurrence in this sample might be attributable to C. p. dorsalis ancestry. Plastral figures, however, possibly become fainter with age as all three turtles were large (older?). Masat and Musacchia (1965) found four electrophoretic patterns in blood serum proteins of C. picta. These patterns were randomly distributed among the turtles regard- less of subspecies. Although this does not dispute Bleakney’s theory it does suggest that a more widespread introgression could have in- fluenced C. p. marginata.

The plastral figure length and width, in com- parison to the respective length and width of the

WHILLANS and CROSSMAN: MIDLAND PAINTED TURTLE, CENTRAL ONTARIO 55

plastron itself, appears to be the most readily quantifiable of the figure characteristics. But, it must be expected that of the subspecific para- meters discussed herein, this will be the most variable as a result of the difficulty in describing the plastral figure.

This study clearly demonstrates that even in the north-central part of its range, C. p. marg- inata displays a diversity of individual traits. The necessity of large samples for correctly des- cribing peculiarities in the midland painted turtle is emphasized. It is probable that even in the best of samples, a statistical population of C. p. marginata will only approach the “typical” and defining disalignment, plastral figure dimensions, and light margin widths.

The Nogies Creek painted turtle sample was normally distributed with respect to carapace length. Individual turtles ranged in size from 103 to 181 mm. Carr (1952) described the catches of a New York study as consisting of animals ranging from 106 mm to 175 mm. He reported the largest turtle on record as being 188 mm Jong. Conant (1975) stated that C. p. marginata usually range in length from 115 to 140, the record length being 187 mm. Thus, the north- central Nogies Creek population falls within the range of carapace lengths that would be ex- pected of C. p. marginata. Figure 4 reveals a difference in the distribution of turtle sizes between the male and female subsamples. This difference is worth noting, but the growth relationships and other factors which may have created it are beyond the scope of the present study.

Spring Activities

That turtles respond predictably to tempera- ture has been indicated by several researchers. Sexton (1959) found that mass emigration of C. p. marginata would be initiated only when the temperature of the inlet was 8°C or higher. Ernst (1972) established the critical maximum temper- atures, and the temperature below which dor- mancy occurred. The number of turtles caught per unit of effort in Nogies Creek oscillated with major temperature changes. Greater catches corresponded with higher temperatures but there was often a lag period between the peaks in temperature and activity. Some authors, how- ever, have noted an acute ability to sense light.

56 THE CANADIAN FIELD-NATURALIST

Noble and Breslau (1938) found hatchlings were attracted to high illumination. Ortleb and Sexton (1964) discovered that C. picta not only responded to light but were able to discriminate between light intensities of 0.1 foot candles. Light and temperature in the natural setting are usually closely linked and their respective effects may be confused.

Legler (1954) reported that C. p. be//ii usually nested between 1700 and 1800 hours. Mahmoud (1968) observed nesting between 0500 and 0900 hours and later between 1600 and 2300 hours. The female C. p. marginata at Nogies Creek commenced daily nesting activities probably no earlier than 1400 hours and ceased leaving the water by about 1800 hours. This did not vary noticeably with air temperature. The annual initiation of nesting was also precise. In 1975 it was only 6 days later than in 1974. But 1975 had a remarkably warmer spring and if temperature is instrumental in inducing nesting behavior there should have been an appropriate early start of nesting.

The uniformity of timing in daily and yearly activity would perhaps be more correctly attrib- uted to light than temperature. Sensitivity to photoperiod would adequately explain the tim- ing accuracy noted in this study. It would provide a much more dependable diurnal and annual clue than temperature. By having a set, safe activity period the turtles would not be fooled by short-term temperature changes when a long-term endeavor such as a summer’s incubation is at stake. It is possible that un- favorable temperature may discourage a photo- periodically induced turtle if there is potential danger to the individual or activity. Nogies ’ Creek C. p. marginata were also observed emerging in abnormal abundance following an afternoon shower, possibly to take advantage of the better digging conditions. Other environ- mental factors may occasionally override the basic photoperiodic clue.

The long distance travelled to an “ideal” nest site by some Nogies Creek turtles and the high incidence of turtles found along several exposed sand banks suggests a form of homing. Further- more, marking turtles at aquatic capture sites revealed females as static during this their anticipated active time of the year. The ratio of males to females taken in nets favored males.

Vol. 91

From recapture information, males are clearly more active than the females. The absence of females from the records may be normal if many female turtles were on or near shore. But in their movements to and from nesting sites they should have turned up in the netting results. This can be explained if the females are able to migrate toa chosen nesting site and return to the site of normal activity. A random net distribution would likely be sensitive to a randomly wander- ing population of female turtles yet could con- ceivably be ineffective against a directed move- ment.

Many authors have documented homing in Chrysemys picta: Cagle (1944), Williams (1952), Gould (1959), Ortleb and Sexton 1964), Emlen (1969), and Ernst (1970b). Emlen (1969) pre- sented a strong argument for visual recognition of local topographic landmarks in orientation of C. picta. He pointed out Casteel’s (1911) findings of good visual discrimination in the species as well as a notable long-term memory.

The long-distance travel to appropriate and hidden nesting sites described in this paper may also indicate homing, perhaps as has been hypothesized for marine turtles in the form of a return to the place of birth (Carr 1972) or as a return to a previously discovered suitable nest- ing location.

Acknowledgments

This research study is the result of secondary activities in the Muskellunge Research Project at Nogies Creek Research Station near Bobcay- geon, Ontario. We are grateful to the Canadian National Sportsmen’s Show, the Ontario Ministry of Natural Resources, and Grant A- 1705 National Research Council of Canada for financial support. We also extend our appre- ciation to Fergus McNeil who assisted in collecting the turtles, to Sophie Poray-Swinarski for constructing the figures, to F. R. Cook for his encouragement, and to J. P. Bogart for his suggestions and later criticism of the manu- script.

Literature Cited

Babcock, H.L. 1919. The turtles of New England. Me- moirs of the Boston Society of Natural History 8: 323-431.

Bishop, S.C. and F. J. W. Schmidt. 1931. The painted turtles of the genus Chrysemys. Field Museum of Natural

1977

History Publications, Zoological Series 18: 123-139.

Bleakney, J.S. 1958. Postglacial dispersal of the turtle Chrysemys picta. Herpetologica 14(2): 101-104.

Brown, D.M. 1972. Heat units for corn in Southern Ontario. Ontario Ministry Agriculture and Food Fact- sheet. 4 pp.

Cagle, F. R. 1939. A system of marking turtles for future identification. Copeia 1939(3): 170-173.

Cagle, F.R. 1944. Home range, homing behavior, and migration in turtles. Miscellaneous Publication of the Museum of Zoology, University of Michigan 61: 1-34.

Carr, A. 1952. Handbook of turtles. Cornell University Press, Ithaca, New York. 542 pp.

Carr, A. 1972. Great reptiles, great enigmas. Audubon 74(2): 24-25.

Casteel, D.B. 1911. The discriminative ability of the painted turtle. Journal of Animal Behavior 1(1): 1-28. Conant, R. 1975. A field guide to reptiles and amphibians. Houghton Mifflin Company, Boston, Massachusetts.

- 429 pp.

Crossman, E. J. 1956. Growth, mortality and movements of a sanctuary population of maskinonge (Esox mas- quinongy Mitchill). Journal of the Fisheries Research Board of Canada 13(5): 599-612.

Emlen, S. T. 1969. Homing ability and orientation in the painted turtle Chrysemys picta marginata. Behavior 33: 58-76.

Ernst, C.H. 1970a. The status of the painted turtle, Chrysemys picta, in Tennessee and Kentucky. Journal of Herpetology 4(1-2): 39-45.

Ernst, C. H. 1970b. Homing ability in the painted turtle, Chrysemys picta (Schneider). Herpetologica 26(4): 399-403.

Ernst, C. H. 1971a. Growth of the painted turtle, Chrys- emys picta, in southeastern Pennsylvania. Herpetologica 27(2): 135-141.

Ernst, C. H. 1971b. Observations of the painted turtle, Chrysemys picta. Journal of Herpetology 5(3-4): 216-220.

Ernst, C.H. 1971c. Chrysemys picta. Catalogue of the American amphibians and reptiles. Society for the Study of Amphibians and Reptiles: 106.1-106.4.

Ernst, C. H. 1972. Temperature-activity relationships in the painted turtle Chrysemys picta. Copeia 1972(2): 217-222.

Ernst, C.H. and R.W. Barbour. 1972. Turtles of the

WHILLANS and CROSSMAN: MIDLAND PAINTED TURTLE, CENTRAL ONTARIO 57

United States. University Press of Kentucky, Lexington, Kentucky. 347 pp.

Ernst, C. H. and E. M. Ernst. 1971. The taxonomic status and zoogeography of the painted turtle, Chrysemys picta, in Pennsylvania. Herpetologica 27(4): 390-396.

Gould, E. 1959. Studies on the orientation of turtles. Copeia 1959 (2): 174-176.

Hartman, W.R. 1958. Intergradation between two sub- species of painted turtles, genus Chrysemys. Copeia 1958(4): 261-265.

Hartweg, N. 1944. Spring emergence of painted turtle hatchlings. Copeia 1944(1): 20-22.

Legler, J. M. 1954. Nesting habits of the western painted turtle, Chrysemys picta bellii (Gray). Herpetologica 10(3): 137-144.

Mahmoud, I. Y. 1968. Nesting behavior in the western painted turtle, Chrysemys picta bellii. Herpetologica 24(6): 158-162.

Masat, R. J. and X. J. Musacchia. 1965. Serum protein concentration changes in the turtle, Chrysemys picta. Comparative Biochemical Physiology 16(2): 215-225.

McCombie, A.M. 1959. Some relations between air temperature and the surface water temperatures of lakes. Limnology and Oceanography 4(3): 252-258.

Muir, B.S. 1963. Vital statistics of Esox masquinongy in Nogies Creek, Ontario. I. Tag loss, mortality due to tagging, and the estimate of exploitation. Journal of the Fisheries Research Board of Canada 20(5): 1213-1230.

Noble, G. K.and A. M. Breslau. 1938. The senses involved in the migration of young freshwater turtles after hatching. Journal of Comparative Psychology 25(1): 175-193.

Ortleb, E. P. and O. J. Sexton. 1964. Orientation of the painted turtle, Chrysemys picta. American Midland Naturalist 71(2): 320-334.

Pough, F.H. and M.B. Pough. 1968. The systematic status of painted turtles Chrysemys in the northeastern United States. Copeia 1968(3): 612-618.

Sexton, O. J. 1959. Spatial and temporal movements of the painted turtle, Chrysemys picta marginata (Agassiz). Ecological Monographs 29(2): 113-140.

Williams, J. E. 1952. Homing behavior of the painted turtle and musk turtle in a lake. Copeia 1952(2): 76-82.

Received 23 June 1976 Accepted 21 November 1976

Germination Requirements of Alaskan Rosa acicularis

R. DENSMORE and J. C. ZASADA Institute of Northern Forestry, USDA Forest Service, Fairbanks, Alaska

Densmore, R.and J.C. Zasada. 1977. Germination requirements of Alaskan Rosa acicularis. Canadian Field-Naturalist 91(1): 58-62.

Abstract. Germination requirements of Alaskan seeds of Rosa acicularis, a common shrub in the boreal zone of Asia and North America, were determined from laboratory experiments and observations of germination under outdoor conditions. Germination was rapid and complete at temperatures of 5°C to 20°C after 2 months of warm stratification and 3 months of cold stratification. Cold stratification alone or with a pretreatment of concentrated H,SO,, resulted in slow and incomplete germination. Laboratory and outdoor experiments indicated that most R. acicularis seeds take 2 years to germinate. Seeds develop and mature the first growing season, warm stratify the next growing season, cold stratify the following winter, and germinate in the spring shortly after snowmelt. Suggestions are made as to the overall reproductive strategy of R. acicularis and the role of its germination requirements.

Rosa acicularis Lindl. (prickly rose) occurson and snowshoe hares. Fruits are also eaten by a broad range of sites in the boreal zone of Asia humans and are an excellent source of vitamin and North America (Figure 1). In Alaska this C. Rosaacicularis has potential as a revegetation species is commoninforestsandisabundantasa species for disturbed areas where food for successional species in disturbed areas. The wildlife and aesthetics are important considera- plant is an important food source for many tions. Little information is available in the animals, including microtine rodents, grouse, literature on regeneration from seed in this species; however, Babb (1959) recommended soaking seeds of R. acicularis in concentrated H,SO, for 1 h, then cold-stratifying at 6.5°C for 3 months, but he did not report what work was done as the basis for his recommendations.

Other species of the genus Rosa that have been studied appear to have the same type of dormancy and similar germination require- ments, although depth of dormancy varies. Temperature-zone Rosa species required cold stratification to break dormancy and germ- inated poorly or not at all when kept at warm temperatures (Blundell and Jackson 1971; Crocker and Barton 1931; Nyholm 1955; Semeniuk and Stewart 1966; Svejda 1968). Pre- treatments of soaking in concentrated H,SO, (Blundell and Jackson 1971; Svejda 1968; United States Forest Service 1948) or warm stratification followed by cold stratification (Bouillene-Comhaire 1970; Rowley 1956; Sem- eniuk and Stewart 1966; Svejda 1968) often gave higher and more rapid germination than cold stratification alone. This paper describes the germination requirements of R. acicularis as determined from laboratory experiments and

FIGURE |. Distribution of Rosa acicularis Lindl. is outlined 3 : : : on the map. Seeds for this study were collected near 4 Observations of germination of seeds subjected

Fairbanks, Alaska. to outdoor conditions.

58

1977

Materials and Methods

Hips! were collected in September 1972, May 1973, September 1974, and September 1975 near Fairbanks, Alaska (64°51’N, 147°44’ W). Achenes were separated from the pulp, washed, dried at room temperature for 48 h, and stored at 2-3°C in plastic bags.

Laboratory germination experiments were conducted in controlled-temperature (+1°C) growth chambers with light/dark period of 16/8 h under white fluorescent tubes, light in- tensity 200 to 500 foot candles. Distilled water was used for all laboratory experiments. Seeds were germinated on vermiculite in plastic con- tainers with transparent lids and three or four 50-seed replications were used for each treat- ment. In laboratory tests, emergence of the radicle was regarded as germination, but seed- lings were transplanted to soil to check for normal development.

To determine the conditions necessary to break dormancy, the following laboratory ex- periments were conducted on seeds collected in September 1974.

(1) Seeds were cold stratified at 5°C for | year. Germination occurred at stratification tempera- tures.

(2) Seeds were soaked for | h in concentrated H,SO,, rinsed, and stratified at S°C. The acid treatment reduced the thickness of the pericarp by half. On a