(navigation image)
Home American Libraries | Canadian Libraries | Universal Library | Community Texts | Project Gutenberg | Children's Library | Biodiversity Heritage Library | Additional Collections
Search: Advanced Search
Anonymous User (login or join us)
Upload
See other formats

Full text of "Resource Management Report November 1, 1961"

No. 60 November, 1961 




ONTARIO 



RESOURCE MANAGEMENT REPORT 
(Formerly Fish and Wildlife Management Report) 



PROVINCE OF ONTARIO 
DEPARTMENT OF LANDS AND FORESTS 

Fish and Wildlife Branch 

(These Reports are for Infra-Departmental Information 
and Not for Publication) 



Hon. J. W. Spoener F. A. MacDougall 

Minister Deputy Minister 



RESOURCE MANAGEMENT REPORT 
(Formerly Fish and Wildlife Management Report) 

TABLE OF CONTENTS 
No. 60 November, 1961 



Page 

Michipicoten Island Beaver Trapping, 1960-61. 

- by E. A. Pozzo 1 

Canada Geese Banding - Lake St. Lawrence, 1961. 

- by J. B. Dawson 4 

Appraisal of Duck Nesting Boxes in Ontario, 1961 - 

A Preliminary Report. - by A. T. Cringan and 

M. G. Johnson $ 

An Analysis of Some Factors Affecting Aerial Moose 

Censusing. - by J. A. Macfie 13 

Woodland Caribou Project - Summer, I960. 

- by D. W. Simkin 23 

The Warm Water Sports Fishery of McGregor Bay, I960. 

- by H. R. Silva and D. I. Gillespie 33 



(THESE REPORTS ARE FOR INTRA-DEPARTMENTAL 
INFORMATION AND NOT FOR PUBLICATION) 



- 1 - 

MICHIPICOTEN ISLAND BEAVER 
TRAPPING 1960-61 

by 
E. A« Pozzo 



Abstract 

A high beaver population exists on Mi c hip i cot en Island 
in Lake Superior due to excellent habitat conditions 
and the absence of predators.. During a beaver census 
in 1959 it was estimated that there were 3500 live 
houses on the islando Up until I960, the amount of 
beaver trapping was negligible.. For the 1960-61 
season, two Indian families from Moose Factory were 
moved to the island by the Department of Lands & 
Forests through co-operation with the Federal Indian 
Affairs Branch. One white trapper also trapped during 
this period. As a result 129$ beavers were trapped. 
The Indian trappers spent some 200 days trapping for 
an average catch of 5°9 beavers per day. The 1176 
pelts flown off the island for the Indian trappers 
brought an average price of $11.91 P er P^t* The 
season's trapping activities appeared to have little 
or no effect on the total beaver population of the 
island. Plans for the 1961-62 trapping season are out- 
lined. 



DESCRIPTION OF MICHIPICOTEN ISLAND 

Michipicoten Island is situated approximately nine miles 
from the nearest point on the mainland and forty-five miles west of 
Michipicoten Harbour. The total area of the Island is 45>344 acres, 
or 70.85 square miles; of this, 1952 acres are covered by water. As 
the Island area is swept by the winds of Lake Superior, it is very 
inaccessible from the mainland by small water craft, and air travel 
is the most suitable means of reaching the Island. 

Due to the great amount of rain, the dense overstory of 
trees, underbrush and the abundant blowdown material, the soil on 
the Island is very damp at all times. It has been noticed that many 
trees have been struck by lightning, but due to the dampness no fires 
originated from these "strikes". The following timber species are 
present on Michipicoten Island: White spruce, balsam fir, cedar, hard 
maple, yellow birch, white birch, poplar and ash. Ground hemlock, 
willow, alder and hazel are also to be found. 

Trappers are the only inhabitants on the Island during the 
winter months. During the summer months a few commercial fishermen 
stay on the Island at the Fishery at Quebec Harbour. Lighthouse 
keepers, a few tourists and fishermen are the only other occasional in- 
habitants. Commercial fishermen and lighthouse keepers leave the 
Island usually in the early part of December. 



- 2 - 

Animal Life 

Beavers are in complete control of the Island. During 1959 
a live beaver house count was made by a white trapper on the Island 
who estimated that there were 3500 live houses. Assuming that this 
estimate was close, the population of beaver on the Island, employing 
an average of three beavers per house, would be approximately, 10,500 
animals, plus bank beavers. Breaking this down further, the population 
would be one beaver for every 4<>32 acres, or 148.4 beavers per square 
mile. The high beaver population might be attributed to the remarkable 
advance in growth of the flora on the Island compared to the mainland 
and the absence of predators. 

No otters, minks, martens, fishers, lynx or wolves are to 
be found on the Island, but there are a few foxes, muskrats, weasels 
and deer. 

The waters on the Island support some very good angling 
for speckled trout, rainbow trout, and lake trout. 

Trapping 

The Island was known as a "poachers haven" for many years 
and trapping signs and old cabins can still be found. Enforcement 
regarding this poaching was difficult because of the relative in - 
accessibility of the Island. However, there is no problem to-day 
due to the close co-operation with trappers placed on the Island and, 
as well, aircraft are available to cope with any enforcement matter 
which might arise. 

It was decided that a trapper be assigned this area during 
the 1956-57 trapping season. However, the trapper 7 s activities did 
not amount to much up until the 1960-61 season, with annual harvests 
as follows: 

1956-57 - 12 beavers 

1957-58 - 45 beavers 

1958-59 - 43 beavers 

1959-60 - 22 beavers 

For the 1960-61 season two Indian families from Moose 
Factory were placed on the Island through the co-operation of the 
Department of Lands and Forests and the Federal Indian Affairs Branch. 
One white trapper was also trapping during this same season. The 
total take for the 1960-61 trapping season was 129$ beavers, 117 6 
taken by the Indian trappers and 122 by the white trapper. 

Camp for the Indian families from Moose Factory was set up 
on Channel Lake on October 7th. On October 11th, a two way radio 
(P-35, Base Portable) was installed at the campsite; so as to have 
communications with White River in case of emergency, to inform the 
District of numbers of beavers taken, aircraft requirements for sup- 
plies and making pickup of fur. Radio communications throughout the 
winter were very good. 



- 3 - 

Following are the dates and amount of fur flown off the 
Island for the Indian trappers . 

January 2nd/6l - 375 pelts 

January 13th/6l - 6£ pelts 

February 7th/6l - 120 pelts 

March 6t;h/6l - 170 pelts 

March 30th/6l - 220 pelts 

May 10th/6l - 223 pelts 



TOTAL 1176 pelts 

This count would have been higher had E„ Trapper not lost 
24 days due to his father's passing away at Moose Factory,, The 
majority of beaver taken were by the triple snare method,, The larger 
percentage of the pelts were very heavily scarred due to the fighting 
among the animals caused by the heavy population pressure . 

The two Indian trappers spent in the neighborhood of 200 
days trapping, this would give an average catch of 5«9 beavers per 
day. The catch shows that one beaver was taken for every 3&»5 acres, 
or 16.6 beaver per square mile„ The trapping activities of the 
trappers in taking this amount of beaver did not seem to alter the 
population too extensively, as the trappers reported that beaver 
seemed as plentiful in the spring as they had been in the fall. 

Following are pelt sizes of beaver taken: 



XXL 


- 


311 


XL 


- 


234 


L 


- 


233 


LM 


- 


101 


M 


- 


121 


Sm 


- 


101 


itts 


- 


25 



The average price per pelt received was$Ll„91. 
Plans for 1961-62 

It is our intention to set up a project during the 1961-62 
trapping season to enable us to gather data pertaining to population 
density, an assessment of food utilization, age-class distribution 
and length-weight-age relationships, as well as trapping effort and 
success. An aerial live beaver house count will be undertaken during 
October. It is also our intention to obtain two more families of 
Indians from Moose Factory, thus giving us four families on the Island. 
This will no doubt give a better coverage of the Island and also will 
reduce the beaver population to a more desirable level. 

Acknowledgements 

J. L. Grew, Fur Management Supervisor, for his assistance 
and guidance throughout the trapping on the Island and Conservation 
Officer H. W. McCullough for his assistance. 



- 4 - 

CANADA GEESE BANDING - 
LAKE ST. LAWRENCE, 1961 



by 
J, B. Dawson 



Abstract 

Canada Geese have nested along the St. Lawrence River 
in the Lake St. Lawrence area for the second year in 
succession. At least 32 young are believed to have 
been produced from ? wild ? nestings. The complete 
flock of 61 birds, adults and young, were captured 
while in flightless condition and banded. 



For the second year in succession, a good number of geese 
have nested along the St. Lawrence. After the broods hatch, they 
congregate together and, at one time, I personally have seen twenty- 
three young, comprising five broods. The caretaker at Nairn Island 
states that he has seen 32 young at one time, and since his observa- 
tions have been reliable, it is probable that at least this number 
of young have resulted from "wild" nestings. Sixteen young were 
produced by captive birds. 

This year efforts were made to capture both young and 
adults while in flightless condition. This was extremely easy to do, 
and the complete flock of 61 birds were driven into a snow fence 
enclosure, checked, and banded, by Gawley and McLeod. 

25 of these were juveniles. 
"i§ were New York banded adults. 
3 were from the flock of 55 Delaware birds and 7 

Codrington birds released in the spring of I960. 
9 were unhanded adults which may have been two year 
olds (at least 24 young were reared outside the 
enclosure last year, but these were not banded) . 
9 were adults received from Quebec and banded by us 

on July 21 (released free-winged in front of enclosure) . 



"7 



The fifty-five Delaware birds and seven Codrington birds 
were banded using the series: 

537-77501 to 5^7-77562 
which were sent from Maple on July 2&, 1959. 

This year we banded all birds using bands supplied by 
Mr. Hugh Schultz of the Canadian Wildlife Service, Ottawa, and under 
his permit number. 



- 5 - 

APPRAISAL OF DUCK NESTING BOXES IN 
ONTARIO, 1961 - A PRELIMINARY REPORT 



by 
A. T. Cringan and M. G. Johnson 
Project Leaders 



Abstract 

An appraisal program was started in 1961 to obtain 
information about factors which influence the effect- 
iveness of duck nesting boxes put out by Sportsmen* s 
Clubs, Conservation Authorities and other organiza- 
tions. Two hundred and twenty-three nesting boxes 
in four separate areas of southern Ontario were in- 
spected during May and June. It was ascertained that 
124 or 59o6% were in or had been in use by wildlife* 
No Wood buck nest was found but six boxes were used by 
Hooded Mergansers while Starlings utilized 74 or 35. 6%. 
Detailed tables showing results are presented. Four 
important factors to be considered in any nesting box 
program are recognized by the authors, namely; the 
nesting tradition of ducks; disturbance by man; competi- 
tion with Starlings and predation. 



Background 

The Wood Duck, considered by many people to be our most 
beautiful native duck, was scarce fifty years ago, and seemed to be 
faced with the possibility of extermination. The reasons for its 
scarcity were complex and poorly understood, but there is little 
doubt that one important factor had been the clearing of the forests, 
destroying most of the large trees with cavities that were suitable 
as nesting sites for Wood Ducks and other tree-nesting ducks. 

The Wood Duck has increased in recent years, and is no 
longer an endangered species. One reason for this increase has been 
the placing out of large numbers of artificial nesting boxes in many 
parts of the Wood Duck T s nesting range, which have been effective in 
allowing small local populations to build up and spread out. 

Some nesting boxes have been placed out in Ontario during 
the past twenty-five years, but little is known of their effectiveness 
in attracting Wood Ducks and Hooded Mergansers here. An appraisal 
program, designed to produce information about factors which influence 
the effectiveness of nesting boxes, and in which Sportsmen* s Clubs, 
Conservation Authorities and other organizations are participating, 
was started in 1961. Two hundred and twenty-three nesting boxes in 
four different areas of southern Ontario were inspected during May and 
June. Most of these boxes had been placed out within the previous 
four months, although some had been out since 1958 • 



- 6 - 

Results 

Some 208 of the structures, or 93% were functional,, That 
is, they could have been used by ducks. Of these, 124 or 59° 5% 
either were in use or recently had been used by wildlife. No Wood 
Duck nest was found, but six boxes were used by Hooded Mergansers. 
Other birds, mostly Starlings, used 10$ of the boxes, and 1$ boxes 
showed use by mammals, mostly squirrels and raccoons. Detailed tables 
showing results are appended to this report. 

The success of the six Hooded Mergansers nests was low. 
Three had been predated and two deserted prior to the time of inspec- 
tion. One had a chance of being successful. 

The objective of this program is to assessthe factors that 
influence the success of duck nesting boxes. Insofar as no Wood 
Duck nests and few Hooded Merganser nests were found, it is not yet 
possible to lay down firm rules that are highly likely to lead to 
success in this type of management. Nevertheless, four important 
factors which should be considered in any nesting box program were 
recognized: 

1. The nesting tradition of ducks is a factor. Ducks tend 

to return to where they were reared for their own nesting, 
and often select the same type of nesting site in which 
they were hatched. Nesting boxes are more likely to be 
successful if tree-nesting ducks already are established 
as nesting species. Boxes are more likely to be used if 
there is a scarcity of natural cavities suitable for nests 
rather than if there is a surplus of such cavities. 

2. Disturbance by man is a factor. Other things being equal, 
tree-nesting ducks will select remote, seldom-disturbed 
places for nesting. Nesting boxes put out in such places 
have a better chance of succeeding than those in sites 
subjected to much disturbance. 

3. Predation is a factor. Entrances to nest boxes should 

be kept small, to reduce access by raccoons. Boxes should 
be deep, so that raccoons will be unable to reach in and 
kill incubating ducks. In many cases, it is worthwhile 
to wrap sheet metal around trees and posts on which boxes 
are placed. 

4» Competition with Starlings is a factor. Many of the 

functional nesting boxes in good sites were occupied by 
Starlings. Apparently Starlings caused two of the six 
duck nests found to be deserted. Reports from Illinois, 
Pennsylvania, New Hampshire and Massachusetts agree that 
competition with Starlings is an increasingly important 
factor that affects the use of nesting boxes by ducks. 
Limited success has been achieved in certain states by 
patrolling nesting boxes every week or so during the 
nesting season and breaking up all Starlings' nests. 



- 7 - 

Recommendations 

Briefly, best results from nesting boxes can be expected 
if predator-proof boxes are placed out in undisturbed places where 
Wood Ducks are known to nest and where there are few unoccupied 
natural cavities, and if the boxes are inspected regularly to prevent 
their being taken over by Starlings,, Few of the nesting boxes inspect- 
ed this year met all of these requirements. 

It is felt that this study should be extended by one more 
year, so as to test further the validity of these first conclusions, 
which testing is necessary to achieve the primary objective of the 
appraisal program, namely assessment of those factors that influence 
the effectiveness of nesting boxes in Ontario. 

The problems of disturbance, predation and competition 
are imposing ones. It will be necessary to move many nesting boxes, 
apply predator guards and make frequent inspections. Probably fewer 
boxes, placed out with greater care, will yield better results. 
Co-operators will be contacted by early winter, in order to prepare 
for the changes that are required,, 

It is unnecessary for co-operators to check nesting boxes 
again this summer, because of the low rate of use by ducks and because 
of the need to move boxes during the winter. It is planned to re- 
check those boxes which contained duck nests, so as to establish the 
outcome of each nest. 

Acknowledgments 

Members of the Simpson* s-Sears Rod and Gun Club, together 
with personnel of the Metropolitan Toronto and Region Conservation 
Authority, who have cooperated in a Wood Duck nesting box program 
since 195$ , initiated the drive that led to this broader program of 
appraisal. Members of the Norfolk Fish and Game Protective Associa- 
tion and personnel of the Big Creek Conservation Authority cooperated, 
again expanding a program already started independently. Personnel 
and members of the Otonabee and Region Conservation Authority, along 
with members of the Peterborough Nature Club participated in that 
region. Members of the Dufferin-Northern Peel Anglers' and Hunters 9 
Association and personnel of the Grand Valley Conservation Authority 
cooperated in the Upper-Grand River Valley area. The assistance of 
these organizations and their many members, in building and setting 
out nesting boxes and in sponsoring the appraisal is appreciated 
greatly. 

The Ontario Department of Lands and Forests has aided the 
study by the collection of background information from throughout 
southern Ontario. Certain administrative services, essential to the 
conduct of any program involving the cooperation of as many individuals 
as this one, have been provided by the Metropolitan Toronto and Region 
Conservation Authority, the Ontario Agricultural College, and the 
Ontario Waterfowl Research Foundation. 

The aid and advice of the many waterfowl biologists in the 
U.S.A., freely given both in publication and by private correspondence, 
who have many years' experience with Wood Ducks behind them, has been 
invaluable. 



- 8 - 

WOOD DUCK NESTING BOX APPRAISAL - GRAND SUMMARY 

Total nesting boxes available for inspection 

223 



Nesting boxes inspected 

Functional 20£ 

Not functional 15 s 

Nesting boxes not inspected 

Seen, could not be checked 4 

Could not be found $ 

Still to be visited 25 



37 



260 



Usage of 208 functional nesting boxes by wildlife 

Number 
In use by wildlife 124 * 

Not in use by wildlife $4 



Used by Wood Ducks in 1961 

Used by Hooded Mergansers in 1961 

Used by Starlings in 1961 

Used by Tree Swallows in 1961 

Used by unknown small bird in 1961 - 

(Probably mostly Starlings) 
Used by Blue Jay in 1961 
Used by Woodpecker, 1960-61 

Used by squirrels, 1960-61 

In use by raccoon 

Used by raccoon, 1960-61 

User not recorded probably Starling) 

Used by unknown duck nest predator 





6 

74 

13 

16 

1 
1 

9 
1 
5 

1 

1 



Per cent 


59.6/o 


40 c 4% 


0% 

2 9% 

35.6$ 

6.2$ 

1.1% 


0.5$ 
0.5$ 


4.3$ 
0.5$ 
2»4$ 


0.5$ 


0.5$ 



3€ Non-functional boxes include one used by squirrel and one 
used by raptore during 1960-61. 

Four boxes were used by two species, hence specific usages 
total 128 rather than 124. 



- 9 - 



WOOD DUCK NESTING BOX APPRAISAL - BIG CREEK REGION 



Total nesting boxes available for inspection 



50 



Nesting boxes inspected 
Functional 
Not functional 


27 
5 


32 




Nesting boxes not inspected 
Seen but could not check 
Not yet visited 


1 
17 


18 




Usage of 27 functional nesting boxes by 


wildlife 








Number 


Per cent 


In use by wildlife 




15 


55/* 


Used by Starlings 

Used by Tree Swallows 

Used by unknown small birds 


(probat 


5 
1 

)ly 9 


1S% 
33$ 



mostly Starlings) 
Not in use by wildlife 



12 



K5% 



Comment 



Use of nest boxes by Starlings and disturbance by people 
are the serious problems in this area. A few Wood Ducks nest in 
the watershed, and careful placing of boxes should lead to success. 



- 10 - 

WOOD DUCK NESTING BOX APPRAISAL - UPPER GRAND RIVER VALLEY 

Total nesting boxes available for inspection 101 

Nesting boxes inspected 92 

Functional 90 

Not functional 2 

Nesting boxes not yet inspected 9 

Seen but could not be checked 2 
Not found 7 

Usage of 90 functional nesting boxes by wildlife 

Number Per cent 

Not in use by wildlife 21 23$ 

In use by wildlife 69 11% 

Used by Starlings 55 61% 

Used by Tree Swallows § 9%> 

Used by unknown bird 4 5% 

Used, but user uncertain 1 1% 

Used by Raccoon 1 \% 



Comment 

About half of the boxes in this sample had entrances which 
were too large. Use of boxes by Starlings seems especially heavy in 
this area,, Most boxes will have to be re-located. 



- 11 - 



WOOD DUCK NESTING BOX APPRAISAL - METROPOLITAN TORONTO REGION 

Total nesting boxes available for inspection 57 

Nesting boxes inspected 49 

Functional 46 

Not functional 3 x 

Nesting boxes not yet inspected 8 

Usage of 46 functional boxes by wildlife 

Number Per cent 
Not in use by wildlife 16 35% 

In use by wildlife 30 



Used by Wood Duck in 1961 0% 

Used by Hooded Merganser in 1961 5 11% 

By Hooded Merganser * Starling 2 

By Hooded Merganser + Raccoon 1 

By Hooded Merganser + unknown nest 

predator 1 

By Hooded Merganser alone at time 

of check 1 

Used by Squirrels during 1960-61 8 17% 

Used by Raccoons during 1960-61 5 11% 

Used by Starlings during 1961 11 28% 

By Hooded Merganser + Starling 2 

By Starling alone 11 

Used by Woodpecker during 1960-61 1 2% 

Used by unidentified small bird in 1961 1 2% 

One non-functional box had been used by squirrels and one by 
a Screech Owl or Sparrow Hawk during 1960-61 «» 



Comment 



This sample includes several excellent sites, including 
several with a history of nesting by Wood Ducks or Hooded Mergansers 
in the past, but not used in 1961. Efforts in this area should be 
to make those boxes already in good sites more predator-proof, and to 
re-locate those boxes in places now subject to disturbance. 



- 12 - 



WOOD DUCK NESTING BOX APPRAISAL - OTQNABEE RIVER REGION 

Total nesting boxes available for inspection 52 

Nesting boxes inspected 50 

Functional 45 

Not functional 5 

Nesting boxes not inspected 2 

Seen but could not be 
checked 1 

Not found 1 

Usage of 45 functional nesting boxes by wildlife 

Number Per cent 

Not in use by wildlife 35 7&% 

In use by wildlife 10 22$ 



Used by Hooded Merganser 1 2% 

Used by Tree Swallow 4 

Used by unidentified bird 2 

Used by Starling 1 

Used by Blue Jay 1 

Used by Squirrel 1 



Comment 

There appeared to be many natural nesting sites in this 
region. Nesting boxes are not likely to be highly successful until 
natural nests produce surpluses of birds that are not able to locate 
natural cavities as nesting sites when they return a year later. 



- 13 - 

AN ANALYSIS OF SOME FACTORS AFFECTING 
AERIAL MOOSE CENSUSING 



by 
J. A. Macfie 



Abstract 

Experimental work to determine the relative success 
of aerial moose censusing under various conditions 
was conducted in Geraldton District during the winter 
of 1959/60. Between December 7 and April 1, IS 
flights were made over a standard 25 square mile 
moose census plot and the moose counted,, Conditions 
that prevailed at the time were recorded such as 
temperature, wind direction, snow depth, days since 
useful snowfall, snow on trees and ground objects, 
sunshine, and groups found by track search. It is 
concluded that for the most accurate results moose 
censusing should be done only when evergreens and 
ground objects are well coated with snow, or on a 
sunny day, when an inch or more of snow has fallen 
24 to 4$ hours before the flight,. 



Considerable effort is being expended in the use of the 
Gogama aerial moose counting method for moose inventory purposes. 
Many conditions influence the success of an aerial moose census; 
all reporters on the subject agree that there are good days for 
counting moose, and there are poor days. If the results of a 
census are to be trusted it must be done under conditions which 
are as nearly ideal as possible. 

During the winter of 1959/60 experimental work involving 
moose censusing was conducted in the Geraldton District. One of 
the objectives was to determine the relative success of censusing 
under various conditions. 

Method 

Between December 7 and April 1, eighteen flights were 
made over a standard twenty-five square mile moose census plot. 
For various reasons it was not possible to give equal time to all 
kinds of flyable weather. For instance, more flights should have 
been made on cloudy days, and more flying should have been done 
early in the morning and late in the afternoon. The plot was 
situated 15 miles south of Geraldton, in an area which has been 
extensively cut-over for pulpwood. Forty-nine per cent of the 
plot has been cut-over in the last 15 years (this includes a 
section where cutting was going on throughout the time the experi- 
ment was in progress). Twenty-seven per cent was coniferous type, 
most of it in one corner of the plot. Ten per cent was mixed wood, 
three per cent hardwood, five per cent stagnant spruce, and six 
per cent water. Except in the large coniferous stand, observing 
conditions were better than average for this part of the Province. 
The topography was rolling, but not precipitous. 



- 14 - 

Airspeed was usually 75 m.p.h., and the height of the 
aircraft above the ground probably ranged between 550 and 
700 feet. It was not possible to maintain the same crew throughout 
the winter, due to illness of the pilot, and other demands on the 
two observers. Only the navigator took part in all IS flights. 
The regular crew was intact (and occupied the same posts) on ten 
flights. 

The block method of searching, where the plot is divided 
into sub-plots having natural features as boundaries, was used. 
The blocks were always flown in a clockwise direction, and in all 
but one case they were searched in the same order. Most track 
searching was also done in a clockwise direction. 

As the project progressed, the navigator became increas- 
ingly aware of where moose were likely to be found. As far as 
possible this was not permitted to help in finding moose, but it 
undoubtedly did increase some counts, or at least reduce the time 
it took to find some moose. This factor, whatever its effect, 
probably did not change much after the first six flights. The 
observers were not in a position to anticipate moose, and it is 
unlikely that the pilot was influenced, either in his flying or 
discoveries of moose, by previous experience. 

Results 

It would probably be impossible to accurately weigh all 
the variables that affect aerial moose counting. In order to get 
enough data the test plot would have to be censused at least 
once a day throughout the winter, and to eliminate bias in such 
frequent counts, several crews would have to be used, which would 
introduce more variables. In order to establish a basis for 
comparing results a rate of success ("estimated efficiency" in 
Tables I and II) has been established for each flight. In the 
early flights considerable guesswork had to be used in deciding how 
many moose were missed in a census, but as the navigator became 
familiar with the plot and many of the moose on it, it became possible 
to estimate with confidence. It is admitted that estimated efficiency 
is a weak link in the reasoning that follows, but it is the only 
way in which a means of comparison can be obtained. 

In Table I, the census flights are listed in chronological 
order, and the counts of moose obtained are shown. Table II rates 
the eighteen flights in the order of their assumed success, and 
shows the conditions that prevailed during each flight. Due to 
the difficulties it presents, no attempt is made to include crew 
combinations among the conditions that affect a census, except that 
on the ten dates underlined in Table II, the same pilot and crew 
were used. 

Discussion on Table II 

Crew ; There was a noticeable difference in the ability of the 
two pilots to discover moose personally. One discovered an average 
of slightly less than one group of moose per flight, while the other 
found only one-fifth that many, The former also seemed to handle 
the aircraft more to the advantage of the observers in track searches. 
However, the flight considered the most successful of all was flown 



- 15 - 

by the pilot who in some ways seemed to be less competent. This 
suggests that, under excellent conditions such as existed that 
day, the ability of crew members isn*t vitally important . Table 
III compares the number of moose discovered by each member of the 
regular crew during eight of the ten flights when that crew was 
intact. Some of the navigator f s success can be attributed to his 
familiarity with the plot, and some to the fact that he occupied 
the most advantageous position in the aircraft, on the inside of 
the orbit and in front of the right observer. The left observer, 
situated at the poorest observing post, discovered more moose 
than the more advantageously situated right observer. However, 
many moose that the latter would have discovered were seen first 
by the navigator sitting in front of him. It is not possible to 
rate observer efficiency under these conditions. 

Date (Col.l) 

Probably the only significance here is that in late winter 
the sun melts snow off trees and ground objects more quickly, and 
crystallized snow causes glare. These adverse effects begin to 
appear by mid-February in this latitude. 

Time census began (Col. 3) 

See Table IV for relationship between time of day and 
whether moose were standing or lying, and Table V for a comparison 
of moose observation and time of day. Table IV suggests that moose 
are more likely to be lying down (and harder to find) within lg or 
2 hours of noon. The most productive section of the plot was 
censused last (except in one case) therefore it is logical that 
the rate of discovery should be greater toward the end of a census. 
This does not entirely account for the high discovery rate for 
late afternoon shown in Table V. Perhaps it is a good time to 
census moose. 

Wind (Col. 5) 

A strong wind makes censusing risky and uncomfortable. 
Moose might dislike wind, taking shelter in a swamp if necessary, 
where they are hard to see. The study reveals no significance in 
wind direction. 

Snow depth (col. 6) 

Our best counts seem to favour the more extreme snow 
depths. This is no doubt partly due to the more effective masking 
of small trees, upturned roots, etcetera, which darken the back- 
ground in shallower snow. Better visibility of tracks in deep snow 
and a possible reduction in wandering of moose, might be minor aids. 

Days since useful snowfall (Col. 7) 

A useful snowfall is one that dates moose tracks made 
prior to it, which requires a fall of about three-quarters of an 
inch. The four most successful flights (efficiency over 90%) took 
place one or two days after a useful snow, and this is significant. 



- 16 - 

Track searches are time consuming, tiring, and often unsuccessful, 
when tracks three or more days old cannot be easily distinguished 
from fresh ones. Under certain conditions a successful census can 
be done without using tracks , but tracks known to be not more than 
one or two days old are valuable clues at anytime. 

On December 2$, we found that 21+ hours after a storm 
many moose hadn ? t made enough tracks to betray their presence 
Probably an interval of 36 hours is best. 

Snow on trees and ground objects (Col. 8 ) 

This is probably the most important factor in moose 
censusing. When conifers, windfalls and upturned roots have a 
thick covering of snow, and fresh snow clings to shrubs, moose 
stand out in sharp contrast to the all-white background. It is 
on such days that successful censusing can be done under an over- 
cast sky* With distraction further reduced by the absence of 
shadows from evergreens, a cloudy day after a storm might be best 
for censusing moose in dense coniferous stands. Column 8 illustrates 
clearly the relationship between snow on trees and the degree of 
success experienced in a moose census. 

It is worth noting that a coating of hoar frost can, on 
the other hand, be a handicap. One census took place when hard- 
wood twigs were heavily coated with frost. This formed a canopy 
through which moose were hard to see. 

Sunshine (Col. 9) 

Too little flying was done on cloudy days to test that 
condition properly. If there has been no recent snow, and tracks 
are of no value, probably a better job can be done under cloud, 
due to the absence of shadow. On a cloudy day, moose look black, 
while conifers and other ground objects look dark grey. But a 
moose in the shade on a sunny day does not stand out from its 
surroundings. 

Groups found by track search (Col. 10) 

On ten of the eighteen flights no moose were found by 
searching tracks, while on the other eight from 18% to 67% of 
the groups of moose recorded were found in this way. On cloudy 
days, or when several days had elapsed since a useful amount of 
snow fell, tracks were of little or no help. In sunshine, from 
one to three days after a snowfall, tracks could be very useful, 
as they were on the two most successful flights. In contrast, 
the third most successful census was made without searching tracks. 
In this case, two other favourable factors, (heavy snow on trees, 
and extreme cold which forced moose into the sunlight) permitted 
the moose to be found without searching. It is likely, though, 
that in some cases here, as in most censuses, tracks guided the 
eyes of the observers to moose found 'off-hand". No attempt has 
been made to weigh the value of tracks where the course of the 
aircraft was not interrupted to investigate them. But it is safe 
to say that on almost every census tracks were of more help than 
column 10 indicates. 



- 17 - 

Conclusions 

The most important single factor influencing the success 
of these censuses was the degree to which evergreens, shrubs and 
other objects protruding above the snow on the ground, were coated 
with snow, Immediately after a heavy snowfall is the best time for 
counting moose. Ideal conditions seldom last longer than one or 
two days after a storm, for wind, thawing by the sun and evapora- 
tion, soon change the background from white to grey spotted with 
innumerable dark objects, each of which must be investigated by 
the eyes of the observer. With a good snow mantle, censusing was 
successful with or without sunshine. 

Fresh tracks were next in importance. If a useful snow 

fell between 24 and 4$ hours before the flight, and the sun 

was shining, a satisfactory census could be carried out even if 
the snow had not stayed on the trees. 

One successful census (March 7) was carried out under 
circumstances other than the two mentioned above. Under what 
seemed to be very poor observing conditions, moose were discovered 
with ease, probably because they were favouring forest openings 
on that warm, sunny, windless day. 

Undoubtedly, there are other combinations that, like 
the last mentioned, favour moose censusing, but they would be hard 
to define and use. For accurate results, at least on plots similar 
to the one used in this experiment, moose censusing should be done 
only when evergreens and ground objects are well coated with snow, 
or on a sunny day, when an inch or more of snow has fallen 24 to 
4# hours before the flight « The ideal situation would be a combi- 
nation of these two. 



- 18 - 



-4 



u-\ 



cv 



ON rH 

• o 

rH CV 



IT\ 



CV 



O 



cv 



Lf\ 



CV 



9 

H 



ON 



CV 

■ 

CV 



-4" On 

9 

CV (H 



CV 



CV 



CV 



-00 



ON -co 



CN, 


CV 


o 


IT\ 


9 


• 


• 


9 


-4 


UN. 


-co 


-co 


rH 









-4 



en 



-4" O 



0"\ 



-4 <r\ \0 



<r\ 



c<\ 



-d (V vO 



tr\ 



£> ur\ vo 



en 

9 

o 

rH 



a 

-co 



CV vO 

9 • 



o 


en 


O 


O 


-co 


-co 


ON 


ITV 


o 


NO 


to 


0^ 


NO 


o 


O 


CV 


rH 


o 


£>- 


ON 


•to 


ON 


ON 


ON 


CV 


o 


rH 


r-\ 


rH 


rH 


rH 














rH 


rH 



UN 



en 


NO 


to 


CM 


ON 


rH 
rH 


ON 


r- 



-4 



^ ^ ^. 

O O tfN 
-CO NO NO 



^ ^ ^ 

en en cv 

on to -co 



en-4C"--ON-4-ONenr>--4m-C0 
OnutnOnvO O- {>- to -4" C*- On ir\ 



-co 



m 



c- 



cv 



Lf\ m 



cv 



en 



•CO NO tr\ -d- 



LT\ 



to 



t> 



o 
cv 



IT\ 



en 



U"\ 


UTN 


en 


UA 


IT\ 


ON 


-co 


CV 


CV 


CV 


CV 


H 


CV 


rH 



[>- ON ON ON 



o c- 



en 
CV 



-co 
cv 



-4 



O 

cv 



cv 



o 

CD 
Q 



o o o 

d) CD CD 
Q Q Q 



C C G 
cd cd cd 
•-a *t> »-a 



fa 



i/\ 



0» 

fa 



ON O- 



o 
Em 



0) 

fa 



CV 
CV 






ON 

cv 



CD 
fa 



t>- 



CT5 



NO 



U 

s 



-4- 

rH 



CV 



cd 



ON 



-co 
cv 



u u 

CO ft 



CD 
CD 
& 
CO 

CD 

o 

T3 
CD 
U 

O 

p 

CO 
ft 

H 

CD 
O 



P 

O 

C 

T3 
•H 
Td 

-P 

o 

H 
•H 
ft 

* 
CD 
S3 



CD 

e 

•H 
P 

<H 

O 
o 

CO 

H 

O 
-P 

CD 

XS 

-d 

CD 
CO 

to 
C 

CD 
O 

■P 

<H 
CD 
rH 

CO 

cd 

p 
o 
H 

ft 

CD 
,4 
P 

<H 
O 

c 

o 

•H 

-P 
O 

CD 
CO 

CD 
> 
•H 
P 
O 

n 
o 

In 

ft 

3 

q 
<3 



H 

H 
aq 

H 



DC 
O 

& 

S W 
13 CO 

o 

fa« 
O 

CO <3J 

a, c£ 

O 



- 19 - 



^ 


v^ 




^ 






^ 






^. 


^ 


^ 






^ 


r- 


OA 


O 


O 


O 


o 


to 


O 


o 


c^\ 


C\2 


to 


O 


O 


o- 


-* 


cn 




C\2 






rH 






■4" 


vO 


o~\ 






vO 



o o 



s 

M 

IT 
CO 

& 
CO 



s 




CD 


O CO 


h 


P 


w 


> 


05 


^H 


03 


Jh 


OCtJ 


0) 


a) 


SE-" 


X! 


T3 


CO 




O 

s 



3£ 


fa O 


o s 


S CO 


M 


CO hJ 


& 


CO fa 


>Hfa 


<U CO 


Q^ 



CO CO CO 

a) a) <p 

K*> r 9 ** P*"» 



> 

05 

CD 
X 



CV 



cv 



CO 

CD 



P>> 

> 
05 

a) 

X 



o 



CD 
-P 
cfl 
S-. 

0) 
T3 

o 
E 



O 



CO 

CD 
P>> 



•H 



to 

rH 



CO 

CD 



T3 
CD 

T5 
Jh 
O 

CD 
P 

o 



cv 



CO CO 

O CD CD 



CD 
P 
05 

Jh 

CD 

o 



CM 



rH 
•H 



P 
•H 



rH -* 



CO CO CO CO to 

CD CD CD CD CD O 

P*"* P*** p-*"> P^ P 5 ** C 



P 

•H 



P 

X 
bO 
•H 



to 



P>> 
> 

05 
CD 
X 



p 

X 
bO 
•H 



{> 



•H 
C 



CV 



•H 



U"N 



CO 
CD 



O 



•H 
PI 



•H 



C\2 



CV 



EC 

O l£ 
S fa 
CO Q 



to -j-cvvo rr\ -d- m -± m o 

CVCVOJCVCviCviCVHCVCV 



-4 
C\J 



C\2 



to NO 



-4 
cv 



cv 



cv 



-4- 



















o 




Q 


i/n. 


ITN 


-4 


£> 


to 


CN2 


-4 


rH 


E 


S 


















rH 


H 


s 


fa 


3s 


3§ 


^ 


^ 


^ 


CO 


05 


35 




B 


CO 


3 






co 




O 



o 

CV 



o 

rH 



> fa 
CO S 



to 



fa 



CV 



fa 

s 



vO 

fa 

s 
w 



05 

o 



Oh 

w 

EH 



-4 -J" rH 

CNJ rH 

+ + I 



+ 



cv 

+ 



-4 

+ 



ir\ 



+ + 



rH 

+ 



to 

H 

+ 



LT\ 



+ 



+ + 



cv 

+ 



ur\ 



r>- upv 

CV C^\ 

+ + + 



CO 
CO 

o < 
O 



o 


CV 


£> 


UT\ 


o 


r^\ 


C\i 


-d- 


LT\ 


-4- 


C^\ 


-ci- 


o 


o 


-4- 


rH 


H 


rH 


rH 


rH 



o 



<r\ 



O 



cv to o 

LTN rH rH 

ON O r^ 

O rH rH 



ifN, cv tr\ 

CV O UA 

rr\ -d- O 

r-\ r-{ r-{ 



O 
rH 



NO 
O 
O 



o c-- 



O 



ON 



w 


o 


Eh 


,Q 


<u 


CD 


p 


Cm 



-oc 


-4- 


c\? 







■ 


u 


c 


td 


05 


S 


•"3 



H 


O 


£> 


o- 


O 


OnI 




CV 




CV 


rH 


C\i 


rQ 


C 


fn 


C 


O 


JO 


CD 


05 


05 


05 


CD 


CD 


fXH 


*-3 


S 


■-o 


Q 


Cn 



cv 

CV 

CD 



cv 

o 

u 

05 



£>- 


to 


<r\ 


rH 


rH 


OJ 


cv 




• 


a 





O 


X 


O 


O 


Jh 


CD 


CD 


CD 


P- 


fa 


Q 


Q 


<l 



in no 



Xi 

(D 
fa 



05 



J> 



o 

CD 
O 



>H 


Q O 


fa B 


Eh fa 


<U M 


So 


M M 


EHfa 


CO fa 


fa fa 



YR 


^ 


^ 


^R 


^ 


^ 


^ 


^ 


^ 


^ 


O- 


Lf> 


c\ 


c°v 


m 


C*> 


cv 


O 


On 


-d- 


Os 


ON 


ON 


ON 


to 


to 


to 


to 


C^- 


t> 



-4-ONUTNOtO ^tO-O- 
C>nOnOnO IA i/N 4 4 



to 
o 
o 



- 20 - 



cv 



CV CV I CV 

I 





EH 


o 




O 


JH 




h-H" 






1— 1 






Ph 


CO 

o 

U 
O 


•r> 












? 




CD 


u 




M 


" 






c 


r 




<-j 


cd 


Ph 




H 


o 


H 


S 


E-; 


■ j 


b£) 


**~ h 


_p 


o 


O 


o 


Pn- 


M 

> 


ch 


CD 


< 




X 


s 


n 


-p 




3 


°H 




o 


o 




U 

o 


ro 






rH 






Oi 






-: 






M 




o 


C 




CO 








f^H 


Ph" 


o 




H 


o 




»~^ 




»""> 


►> 




a 


Ph 




. 




r-1 


•3 




Oi 




PQ 


.o 


fn 


O 


b 


o 




L-> 


S> 


|H 




o 


iJJl 




" 


O 


co 


t- 


H 


Ph 


'J3 


Ph 


2 


Q 




o 


CD 




O 


CO 






o 






o 






s 




CD 


cm 




CO 


o 




o 




Ph" 


q 


C 


w 


Sd 


o 


p> 




CO 


pcj 


rH 


•^ 


Ixl 


CTJ 


J : 


CI 


■P 


d 


PQ 


O 




° 


r j 


o 


EH 




o 


Pn 


CO 




H 


ft 


<tj 


1 
i — . 


2 
c 



j cv 



I rH vO 



cv 



O CV to r^ -4- O vO 






rH 



CM CM vO 4 



-4" 
CV 



u-\ 



C~\ tO rH VO ON. CV r^ 



oa 



OA 



CV t^ H 4 



H H 



vO 



-00 



\Q 



■00 



O 

rH 



UA 



0^ 



-4 



r^ 



<r\ 



IT\ 



cv 



tN 



ON 


to 


-4- 


CV 


O 


O 


to 


cv 


CM 




cv 


CV 




cv 


£.; 


rH 




>> 


!>> 






CD 


CD 


S 


?h 


U 






,c 


rO 


Jh 


ctf 


ctf 






6 


B 


cti 


2 


3 


r^-l 


r-C, 


CD 


CD 


3 


£h 


Sh 


O 


'o 


o 


O 


C 


rQ 


r O 


rH 


rH 


CD 


CD 


rd 


CD 


o 


cd 


jg 


Q 


Q 


•-3 


[x, 


E=< 


S 



rH 


rJ 


•H 


<U 


h 


EH 


ft 


o 


•=31 


E 





CO 




Jh 




CD 




> 




rH 




CD 




CO 




rQ 




o 


CO 




P 


o 


,3 


►-> 


bo 


p 


•H 




rH 


!>> 


«H 


^ 


t>- 


i>> 


H 


rH 




CO 


fn 


2 


CD 


o 


rQ 


CD 


s 


S 


CD 


CTJ 


O 


P 


CD 


rH 


n 


S 




e 


T3 


•H 


ti 


CO 


c6 






T3 


£>- 


CD 




rH 


U 


CJ 


CD 


> 


rQ 


o 


s 


o 


CD 


CO 


o 


•H 


CD 


x) 


Q 






CD 


CD 


CO 


& 


O 


■p 


o 




& 


rH 




o 


<H 


<H 


O 


T3 


CO 


CD 


ft 


T3 


3 


rH 


o 


o 


Sh 


o 


hO 


CD 




U 


CD 




T3 


p 


J3 


o 


H 


ti 


o 




CJ 


co 


•H 


crj 




£ 


P 




O 


PI 


c 


O 




•H 


to 


P 


CD 


cd 


o 


S 


■d 


fn 




O 


CD 


cm 


H 


Pi 


rC 


•H 


Cti 




P 


CO 




•H 


0) 


rCl 


rC 


EH 


Eh 



- 21 - 



TABLE IV: PER CENT OF MOOSE LYING DOWN BY QUARTER HOUR PERIODS 



NUMBER OF MOOSE OBSERVED BY QUARTER HOUR PERIODS. 




o 


in 


o 


ID 


o 


LO 


o 


ID 


O 


in 


O 
CO 


in 


o 


in 


o 
en 


m 


o 
o 


in 


o 
en 


tn 


o 
o 


in 


o 
en 


in 


o 
o 










— 


- 


— 


— 


<\l 


(M 


0,1 


OJ 


en 


en 


en 


cn 


<3- 


<* 


<* 


«r 


LO 


in 









- 22 
TABLE V 



A COMPARISON OF TIME OF DAY AND MOOSE OBSERVED 



The number of times a flight was in 
progress in a given quarter hour. 



The number of moose observed in each 
fifteen minute period. 



NO. OF FLIGHTS AND 
5 10 15 20 



No. of Moose 

25 30 35 40 



10 
10 
10 
10 
I! 
II 

II 
II 

12 
12 

12 
12 

13 

13 

13 

13 

14 

14: 

14 

14: 
15: 
15; 
15: 
15: 
16: 



:00 
:I5 
:30 
:45 
:00 
:I5 

:30 
:45 

:00 
:I5 

:30 
:45 
rOO 
:I5 
:30 
:45 
:00 
:I5 
:30 

:45 
:00 

;I5 
:30 
:45 
:00 



V 1 


• ' 


1 


I 


\ 


/ 






\ / 








\ / 








1 / 








\* 








V v 








\ \ 


t 
1 






\ 


1 

• 






\ 


\ 






/ 


\ 






/ 


N 






/ 




\ 




/ 




«""» 




" 


< 

\ 






T 


s 






I 


•« 






~ / 


S 

» 






/ A 


i 






1 *s 








A 








// 








/* 








Vv 








-\ \ 








\ N 








\ V 








\ * 


\ 






\ 


i 






\ 


S" 






- \ ' 








y 








y\ 








V 








\ ^ 






\ 


\„ 






\ 









\ 


y 






V 


' 






\ 








v 


K 






' 


»v 






■ 


v. 


**• _ 




— 7 








/ 






JU# 


/ 








J 






*~ 


/ 








/ 




/ 




/ 




• 










N 






1 


1 


• 
1 


\ 



- 23 - 

WOODLAND CARIBOU PROJECT - SUMMER, I960 

by 
Do W. Simkin 



Abstract 

A preliminary study of caribou range in the Hudson 
Bay Lowlands was carried out from July 14 to August 
11, 1960o Two main types of lichen woodland were 
sampled. The most important being the more mature 
muskeg lichen woodland and the other a subclimax 
lichen woodland located on old beach ridges,. In 
both types availability of ground lichens was very 
higho From limited observations it appears that ground 
lichens are the important winter food of caribou and 
ground lichens along with various herbaceous species 
constitute their summer diet A summer density of 
one caribou per 3-5 square miles was estimated for 
this study area. Beach ridges, windswept lakeshores 
and open bogs are utilized by caribou in the summer 
presumably chiefly to rid themselves of the multi- 
tudes of flies. A list of birds and mammals observed 
is appended to the report. 



Introduction 

During the period from July 14 to Aug. 11, I960, a prelimin- 
ary study of caribou range in the Hudson Bay Lowlands between the 
Shagamu and Winisk Rivers was carried out. The main purpose of this 
project was to become familiar with what is thought to be representa- 
tive woodland caribou, Rangifer tarandus caribou , summer and winter 
habitat and to make general notes on caribou ecology and behaviour 
where possible. 

Itinerary 

The author and an Indian guide, Michel Hunter of Winisk, 
departed from Winisk and travelled by canoe up to Hudson Bay coast 
to the Shagamu River. We then followed the Shagamu River to its 
source at Shagamu Lake, taking five days to make the inland trip. 
A base camp was established at Shagamu Lake (believed to be typical 
of the main caribou wintering area in the lowlands) from July 27 to 
August 3. Daily field trips from this camp were made from 4-$ 
miles into the surrounding area. On August 4, we headed from this 
camp for Winisk by canoe. Using the many potholes and small streams 
in the area, we were able to move fromthe Shagamu to the Mishamattawa, 
to the Winisk watersheds by canoe without having to make too many 
difficult portages. Numerous walking trips were made from camps 
located along the water route followed. Eight days later we arrived 
back at Winisk. 



- 2k - 

Methods 

Representative stands of ground and tree lichens were 
sampled in the different habitat types encountered on this trip 
and the results were tallied in a similar manner to that used in 
1959 (Simkin I960 (a)) Table I. In different cover types support- 
ing stands of ground lichen, trees representing average condition 
were bored and aged using an increment borer. Table II. 

Although no caribou were seen much fresh sign was observed. 
Notes were kept on relative track abundance in a subjective manner. 
Also, where observed, food species utilization by caribou were re- 
corded. 

In addition to these data some additional knowledge of 
caribou behaviour was gained through indirect observation and 
through many discussions with Michel Hunter, who is a very good 
caribou hunter and who has killed many caribou in the area. 

Description of the Area 

This study area was located within Ahti's Subarctic Lichen 
Belt (Ahti &Hepburn, 1961) and within Coombs ? Muskeg and Small Lake 
Zone (Coombs, 1954) • The Winisk and Mishamattawa Rivers formed the 
eastern boundary while the Shagamu River formed the western. It 
lay between the Hudson Bay coast and a line 50 miles inland parallel 
to the coast. 

The area around Shagamu Lake, in which most of our tra- 
velling and field work was carried out, was believed to be 
representative of the winter caribou habitat. This region was 
characterized by its lack of relief and multitudinous shallow lakes 
and ponds. Numerous small creeks in the area effected very poor 
drainage. Much of the land about these lakes was overlain by peat 
varying in depth from one to ten feet. (Hustich's 3rd. type of 
lichen woodland (Hustich 1957)) • In general this supported a very 
meagre growth of black spruce Picea mariana and tamarack Larix sp. 
and a very luxuriant growth of reindeer lichens, principally Cladonia 
alpestris . C. mitis, C. uncialis and C. amaurocraea . Generally 
C. alpestris was by far the most dominant species and apparently 
had been so for some time. One area examined had C. alpestris 
growing to a depth of 23 cm. on top of a decaying layer of C. 
alpestris . 10-13 cm. deep. 

In this area of accumulation of deep peat layers, the 
topography was extremely irregular. This was due, probably, to the 
action of frost on the deep layers of peat. As a result, the area 
was characterized by many hummocks five and six feet high and five 
to 15 feet wide. 

In some cases deep crevices were apparent between hummocks, 
where frost heaves had split the peat in a vertical direction. It 
appeared that once the surface layer of growing material was broken 
it was susceptible to severe wind damage. On more than one occasion, 
we saw instances where the wind had picked up and flipped over solid 
mats of vegetation up to 30 feet in diameter. 



- 25 - 

Several trees were aged in this area and as suspected 
were found to be extremely slow growing (Table II). 

About four miles north of Shagamu Lake an old sand beach 
ridge, which apparently extended for a great distance N.W. and S.E., 
was travelled for several miles. Unlike the low-lying muskeg areas, 
it was covered with peat to a depth of only two or three inches and 
overlain chiefly with growths of lichens, mainly subclimax species • 
White spruce, the species most frequently occurring on this ridge, 
showed a much more rapid growth rate than did the tree species in 
the area previously discussed. 

Similar characteristics were noted forthe ridges visited 
while ascending the Shagamu River and on a ridge visited about 
10 miles S.W. of the Mishamattawa River, These were true lichen 
woodlands. 

Much willow Salix sp., bog birch Betula glandulosa , and 
herbaceous species such as cottongrass Eriophorum sp., bog bean 
Menianthes trifoliata and horsetail Equisetum sp. was found in the 
wet areas between ridges around the thick peat area and around the 
shores of ponds and lakes. 

The small tributary streams in the upper Shagamu and 
Mishamattawa watersheds were very tortuous and bordered by heavy 
growth of willow and alder Alnus sp,, whose overhanging branches 
made travel by canoe quite difficult in many places. 

Navigable rivers and streams were not very numerous within 
the area but by taking advantage of Michel* s knowledge of the 
distribution of ponds and lakes and some of the streams, we were 
able to travel from the Shagamu watershed into the Mishamattawa 
system by canoe without too much difficulty. 

Availability of Lichens 

(a) Ground Lichens 

Nine different ground lichen sample plots in various 
locations were recorded (Table I) . These were located in 
representative areas. 

An important point which was illustrated by these sample 
plots was the very old age of most of the lichen stands in the 
deep peat muskeg area (hereafter called muskeg lichen woodlands) , 
and the relative immaturity of the lichen stands on the beach 
ridges. 

It appeared that the muskeg lichen woodlands due to the 
deep layer of moist and wet peat and comparatively sparse growth 
of very slow growing trees (Table II) was much less flammable than 
the drier beach ridges. In the muskeg lichen woodland, due to the 
lack of soil, the lichens had much less competition and therefore 
probably became established more quickly than on the ridges. Due 
to the low fire hazard (the main factor disturbing normal plant 
succession) , these stands in general had a better chance to survive 



- 26 - 

for many years without disturbance. In these areas Cladonia 
alpestris was usually the dominant lichen and covered from 30-1 
of the surface area D C„ rangiferina , C. mitis , C. uncialis were 
considerably less common. (See Table I for measurement and species 
composition ) 

Due to the irregularity of the surface in this muskeg 
lichen woodland type, it is possible that the ground lichens would 
be quite available during the winter as the snow would likely be 
kept blown off of the hummock tops and would not pack so hard 
anywhere as it would on a smoother surface such as inthe wetter 
more level inter ridge areas. Indeed several places were observed 
where caribou had obviously grazed during the winter in this type. 
Michel voiced the opinion that caribou preferred this type in the 
winter. 

In contrast to the above, the beach ridges and some of 
the higher river banks were covered with only 1-3" of peat on 
which a vigorous growth of reindeer lichens was found. Generally 
the lichen stands were primarily of the more intermediate succession 
types, chiefly, Cladonia mitis , C. uncialis , C. rangiferina and 
C. amaurocraea . C. alpestris was usually present but due to the 
young age of the stand had not attained dominance as in the case 
of the muskeg lichen woodland. 

It was immediately apparent that the trees growing on the 
ridges (predominantly white spruce) had a much more vigorous growth 
rate than those in the deep peat area. (See Table II) . 

It appears that there is probably a fire cycle on these 
better drained areas. I believe this might be about a 30 to 50 
year phenomenon. After this period of time the trees reach a 
more inflammable stage and the lichens become sufficiently thick 
to provide adequate fuel for an extensive fire. This was suggested 
after observing the average size of trees which had been burned on 
a few of the ridges which we investigated. 

Much sign of caribou use of lichen on these ridge sites 
was observed. Probably because of the higher elevation which 
provided an exposure to breezes to reduce the fly nuisance and 
because of the more solid footing, caribou prefer to travel on them. 

A heavily worn trail could be seen on the crest of any 
ridge we visited. This was according to Michel, worn by the feet 
of Indians as well as ungulates and predators in the old days. Now 
they were being used as caribou, wolf and fox highways chiefly. 
Many wolf and fox scats were observed along such ridge trails. 

(b) Tree Lichens 

The muskeg lichen woodlands were the most important 
producers of tree lichens. In most places availability of tree 
lichens was very good on the stunted black spruce and tamarack. 
Ranking second in importance as producers of tree lichens were the 
relatively fast growing spruce along some of the higher stream and 
river banks. 



- 27 - 

Primarily due to their young age, the trees growing on 
the ridges were of little value as bearers of tree lichens. 



sp 



The important palatable tree lichen species were Alectoria 
, Evernia mesomorpha , Usnea sp. and Cet raria cil iaris in 



descending order of abundance, 

Food Habits 

As stated before no caribou were observed in the area al- 
though considerable fresh sign was seen, The following is a list 
of feeding observations made by simply observing plant utilization 
in areas where caribou tracks indicated what species had been using 
them, 

TABLE III - Caribou Feeding Observations 

(a) Winter Feeding 

(i) 20 ft, square area - C. alpestris . C. mitis . C. uncialis 

(ii) C. alpestris , C. rangiferina 

(iii) Co alpestris - extensive grazing 

(iv) Co alpestris , C, mitis , C. uncialis , C. rangiferina , 

15 ft. square patch utilized about 75^° 

(v) C. alpestris and other Cladonias 



square 



(b) Summer Feeding 

(i) 2 species of willow 

(ii) C. alpestris and C. rangiferina 

(iii) Sali x sp. 

(iv) C. alpestris , and C, rangiferina - many 1 ft 
patches. 

(v) C. alpestris and C. rangiferina 

(vi) Salix sp. and Equesetum sp. 

(vii) Andromeda glaucophylla and Betula glandulosa . 

Likely considerable tree lichen is used during the winter 
but grazing on this type of food is quite difficult to determine. 
Nowhere did I see a tree which had been denuded of its tree lichen 
nor on which much utilization was evidenced. I do not believe that 
tree lichens are as important as ground lichens in the winter diet 
of caribou in this area. 

I believe a much greater variety of herbaceous species 
was eaten by caribou but due to the fact that caribou often eat, 
while on the move, nibbling here and there, it was extremely 
difficult to find plants which had been grazed. However, simply 
by observing concentrations of fresh tracks in various areas, I 
was quite certain that much more use was being made of herbaceous 
vegetation than the above table would indicate. Among the species 
suspected to be of importance are Eriophorum sp., Menvanthes trifoliata 
and numerous Carex sp. which were so common in the wetter areas. 



- 2d - 

(c) General 

It appears that the food habits of woodland caribou in 
portion of the Hudson Bay Lowlands at least, are very similar to 
those of the herd studied in the old Woodland Caribou Crown Game 
Preserve. Reindeer lichens of the Genus Cladonia seem to be the 
staple winterfood, while herbaceous vegetation such as willow and 
bog birch leaves and various smaller herbs become important in the 
summer. Ground lichens are used in the summer also but not as 
extensively as in the winter. 

The important difference between the ground lichen stands 
studied in the Sub Arctic Lichen Belt and the Western Rock Region 
is the difference in the stage of development of the ground lichen 
flora of the two areas. In the SALB Cladonia alpestris . the climax 
species, is the predominant species, while in the WRR, subclimax 
species such as C. mitis, C. uncialis, C. amaurocraea , C. rangiferina 
predominate. 

In both areas the availability of both summer and winter 
food is great and it would be hard to imagine scarcity of food alone 
being a limiting factor on caribou numbers in either area . 

(d) Coprpphagy 

While discussing food habits, I would like to include here- 
an observation which my guide made a few winters ago while hunting 
caribou. He had stalked a herd of seven caribou out onto a lake 
and found them standing around in a close group apparently licking 
something on the ice. After shooting at the herd, he went out to 
investigate and found that they had been licking an otter scat. 
He stated that on another occasion caribou, which he had killed, 
had fragments of otter scat within their stomachs. Apparently this 
is a rather common occurrence for many of the Winisk hunters had 
spoken to him of finding otter scats in the stomachs of caribou 
which they had killed. 

It would be interesting to know why they lick and eat otter 
scats. Is there perhaps a trace element deficiency in their diet 
which they satisfy in this manner? 

Abundance and Distribution 

Our aerial surveys of I960 and 1961 leave no doubt that 
the area studied is part of a major caribou wintering area (Simkin, 
I960 and Simkin, 1961.) 

In 1961, we estimated a wintering population of 4»3 square 
miles per caribou in an area close to the present study area. In 
the I960 survey a population density of approximately one caribou 
per five square miles was found near this study area. 

No sampling system designed to give an estimate of caribou 
numbers was used during this project. It was apparent, however, 
that many caribou were in the area. According to previous interviews 
with local hunters, most of the caribou move out close to the coast 
in the summer and remain there until fall. Obviously such was not 



- 29 - 

the case as almost everywhere we travelled we saw fresh caribou 
sign (see map). I estimate that the density in the area might be 
2-5 square miles per caribou during the summer. In a few places 
near Shagamu Lake doe and calf tracks were seen also. Likely caribou 
calve in this inland area. 

In connection with track observations it is worthwhile 
here reporting that a great deal of otter sign was observed in the 
wetter areas. This animal apparently uses regular runways between 
different areas. There is a good possibility that these otter 
runways through the muskeg could be confused with caribou trails 
by aerial observers on summer surveys unless they were made aware 
of the fact that otter trails do exist and are quite numerous. 
Once this was done it should not be too difficult to distinguish 
between the tracks of the two species. The otter tracks are quite 
deep, of uniform width and run in a more direct line across the 
muskeg than do most caribou and moose tracks. 

Caribou Behaviour 

As stated before it was disappointing that no caribou were 
observed while doing fieldwork. Some of the track observations, 
however, did prove useful in illustrating certain aspects of caribou 
behaviour. 

(a) Aggregations 

It was quite apparent that no large size groups of caribou 
were travelling around in the area we worked in. On no occasion did 
we observe tracks of more than two animals travelling together. 
Usually the pair was comprised of a doe and her calf. On one occa- 
sion, we started two adults which Michel believed were stags from 
the shore of a lake. 

Most of the tracks seen were of single animals traversing 
wet bogs, beach ridges, cottongrass marshes and moving around the 
edges of lakes. 

(b) Use of Cover 

Since no animals were seen on our trip and fresh tracks 
were seen almost wherever we went it must be concluded that the 
caribou although fairly numerous were quite wary and made good 
use of protective cover. 

(c) Habitat Preference 

As stated before the natives figure that most of the 
caribou concentrate along the ridges closer to the coast. This 
they say is mainly to reduce the worries caused by the heavy 
infestations of blood sucking Diptera. Although I do not believe 
that there is such a concentration of animals along the coast, 
it does seem reasonable that during the bad fly season, caribou 
would tend to seek out areas where some relief from flies might be 
found. No doubt harassment by flies is one of the reasons why 
caribou are constantly on the move during the summer. Certainly 
enough tracks were seen in exposed areas such as crests of beach 
ridges, windswept lake shores and along edges of open bogs to 
indicate that much use was being made of this type of habitat. 



- 30 - 

(d) Fall Movement 

While travelling on the ridge north of Shagamu Lake, 
Michel told me that he had seen signs of as many as 50 caribou 
travelling along the ridge in October or November. According to 
him the animals congregate in a large open area between Wood 
Creek and Shagamu River and near the coast in September., Then 
later on they move inland and travel some distance along this ridge 
before heading farther southwest. 

It would be interesting to know if the area around 
Wood Creek was a rutting area and if so how many caribou made use 
of it. Probably the best way to determine this would be by use 
of an aircraft, preferably a helicopter in the fall. 

Summary: 

1. A preliminary study of caribou range in the Sub Arctic Lichen 
Belt was carried out from July 14 to Aug. 11, I960. 

2. Ten sample plots, (nine ground and one arboreal) of represent- 
ative lichen stands were recorded. 

3« Increment borings were made of several trees to determine age 
of the stands and growth rates of trees. 

4« Two main types of lichen woodland were sampled. The most 
important being the more mature muskeg lichen woodland and 
the other being usually a subclimax lichen woodland located 
on old beach ridges. 

5. In both types availability of ground lichens was very high. 

6. Tree lichens were much more plentiful in the muskeg lichen 
woodlands. 

7. It is suggested that about a 50 year cycle is in effect on 
the lichen woodlands on well drained beach ridges. Fire is 
the main habitat destroyer causing a new cycle. 

So From limited observations it appeared that ground lichens were 
the important winter food of caribou in this area and that 
ground lichens along with various herbaceous species consti- 
tuted the summer diet of caribou, 

9. Little difference was observed between the food habits of the 
caribou in the Sub Arctic Lichen Belt and those of caribou in 
the Western Rock Region. The only important difference being 
that in general the SALBwas more nearly approaching the 
climax stage than the WRR. 



i t> v 



10. Mention is made of coprophagic behaviour of caribou eating 
otter scats in the winter time. 

11. A summer density of one caribou per 3-5 square miles was 
estimated for this study area. 



- 31 - 

12. The difference between otter and caribou trails in the muskeg 
is pointed out. 

13 • Apparently caribou travel in only small groups during the 
summer time. No more than two caribou were seen per 
aggregation. The doe-calf aggregation was most frequently 
observed. 

14« Caribou make good use of cover in the summer and appear to be 
very wary. 

15. Beach ridges, windswept lakeshores and open bogs are utilized 
by caribou in the summer presumably chiefly to rid themselves 
of the multitudes of flies. 

I60 One suspected rutting area occurs to the north of the study 
area. It is suggested that an aerial survey in September 
might reveal its exact location and the number of animals using 
it. 

17. A list of birds and mammals observed is appended to this report. 



TA3LE II - Trees Aged in Study Area 



Species 


H 


eight 


Dbh. 


Age 


Site 


Black 


spruce-Picea mariana 




15' 


■~i ?v 


35 


muskeg lichen w 


Black 


spruce 




15' 


3»« 


35 


n « 


Black 


spruce 




15' 


3«» 


130 


JV IJ 


Tamarack-Larix laracina 






C»J 


75 


Ji »l 


Tamarack 






3h n 


50 


t? IV 


White 


spruce-Picea glauca 




10 * 


2" 


11 


old beach ridge 


White 


spruce 




25 ? 


6" 


44 


tv l» «« 


White 


spruce 




10 « 


14" 


10 


f , !» »9 


White 


spruce 




20 * 


c?» 


2$ 


?» U Vi 


White 


spruce 


50 


-60* 


134" 


105 


High river bank 



-■32 - 



TABLE I - Lichen Sample Plots 



ltohew wnnm.Awn typf, 


MUSKEG 


Species 


I 


IT TV V 


YT 



Over- 
story 

Under- 
story 



Sb,Tam. Sbo Tarn, 



15' 
Salix 

Vfo 



30% 
10-15' 

Ledum 

Sb 

30f 



Sbo 

1558 

30' 

Ledum 

Sb 

30$ 



Sbo 

I/O 

15' 



Ledum 
Sb 



15? 



Sb. 

1% 

10' 

Sb. 
Tarn 
Ledum 



i 



Ground 
Layer 



Co alpestris 

Co rangiferina 

Co m itis 

C.o degenerens 

Co amaurocraea 

Co sylvatica 

Co uncialis 

Co cornut a 

Steriocolon sp 

Horn Lichens 

Cetraria sp. 

E mpetrum nigrum . 

Vaccinium vitis idea 

Sedges 

Rubu s chamaemorus 

Andromeda 

Vaccinium sp . 

Equi setum 

Bare Humus 

Moss 

Salix 

Parmelia sp. 
Depth of Peat Over Soil 
Height of Lichens 



BO 


50 


20 


60 


30 


15 


15 


50 


35 
5 


2 
30 



2 
/ 

/ 

1 



1 



/ 



30 



1 

/ 

1 



1 
1 

2 



23 cm, 



25 


25 
1 




3 
/ 


3-5' 


k-y 


K-y 


3-10' 
6.5-7 cm. 






- 33 - 



CO 

W 
DC 
O 



Eh 



> 



XJ 


a 


H 


CO 


O 






cO 





•H 


co 


Ih 


h 


o 


• >» 


-p 


,c 


o 


• »- &o 


Q) 


£> mxJ <r\ 


rH 


CO <H r^rH 


«u 



Ph 
CO 

cd 

•H 

c 

Ih 

<d 
> 



a CO 

CO -H 

rH 



CO 

cO 

•H 

H 

<l) 
S 
u 

cO 



* 











CO 






CO 


p 




O 


3 


» 


rH 


•H 


£3 


CO 


3 


,Q 


cd 


CO 


s 


rH 


id 





E-i 


<D 


3 


3 


a 


rH 


• » 


^Q 


Xi 


-P 


cO 


•H 


^O^ 


•H 


CD 


03 


H 


a 


CO H H 


cd 


HH" 


GQ 


W 


w 



H 



o 

Q 
M 

o 

w 

PQ 



H 

r- 
> 



£0^ 

co c\i tr\ 



S 
cO 



,0 u-n^ 
CO rH wo, 



o 

rH 
•HO 



6 






cO 


S 




Eh 


3 




• 0- 


-a 


^ 


£ tr>^ 


CD 


flO 


co cv ir\ 


1-3 


CO cv 





•H 




c 


g 


•H 


3 


O 


T3 


O 


0) 


cO 


►H* 


> 



o 

H 

H 



CO 
CO 
O 



•H 

o 

rH 
■H 
tH" 

H 



X) 



1 >» 


u >* 


^ k 


CD U 


CD O 


-a 


> -P 


C +3 


O CO 


£3 CO 



on <D 



C\i 



%.n 



o ^O 
-4" -4- 



O 

rH 



tr\ 



000 

-4: CM CV 



O 
CV 



^1v ^ 



i^O HO 
rH t>- r-\ 



XJ" 


S 


c 





cO 




CO u"\ 




• 


^ 


CD 


I 


> 


ir> 





a 


5 


on 


<N 




X> 




C 


e 


cO 


e 


CO 





uo 


CD 


1 


> 


LT\ 


O 


• 


s 


ITS 


•-II0J 




H 




X) 




C 


■ 


cO 


s 


CO 





^-00 


CD 


1 


> 


LT\ 


O 


e 


mm 


vO 


■HN 




XJ 




c 




CO 




CO 


• 




6 


s-4 





CD 




> 


ir\ 


O 







^OH 









f : '-i ; 



i i i 



« •' U' :' 



I".. I' 






I 






APPENDIX I - 34 - 

A. List of Birds Seen Winisk River to Shagamu River 

1. Common loon 

2. Arctic loon 

3. Brant geese 

4. Black duck 

5<> American goldeneye 

6. White-winged scoter 

7» Surf scoter 

£. Mergansers sp. American or red-breasted 

9» Ruddy turnstone 

10. Purple sandpiper 

11. Parasitic jaeger 
12 o Herring gull 

13. Arctic tern 

B. Annotated List of Birds Seen Inland 

1. Common loon - very common 

2. Richardson's goose - only one seen - likely non breeder 

3. Canada goose - very common breeding bird 

4. Mallard - scarce nesting species 

5o Black duck - scarce nesting species 

6. Pintail - common 

7. Green-winged teal - common nesting species 

5. White-winged scoter - scarce - 2 broods seen 
9« Surf scoter - scarce - 1 brood seen 

10. American merganser - scarce 

11. Sharp-shinned hawk - common 

12. Red-tailed hawk - uncommon 

13 . Rough-legged hawk - common 

14. Osprey - common - 3 nests seen 

15 • Spruce grouse - common - 2 broods seen 

16. Willow ptarmigan - common - 4 broods seen 

17. Sharp-tailed grouse - common - 3 broods seen 

18. Coot - uncommon - 1 pair seen 
19° Spotted sandpiper - common 

20. Solitary sandpiper - uncommon 

21. Greater yellowlegs - common 

22. Lesser yellowlegs - common 

23. Herring gull - common 

24. Bonapartes gull - uncommon - nesting species 
25 Common tern - uncommon - nesting 

26. Yellow-shafted flicker - common - nesting 

27 o Yellow-bellied flycatcher - common 

28. Tree swallow - common 

29. Gray (Canada) jay - common 

30. Raven - common 

31. Brown-capped chickadee - common 

32. Robin - common 

33 • Northern shrike - uncommon 

34o Magnolia warbler - common 

35. Myrtle warbler - common 

36. Rusty blackbird - common 

37* Cowbird - 1 pair seen on Shagamu River 

38. Pine grosbeak - common 



- 35 - 

39« White-winged crossbill - common 

40. Tree sparrow - uncommon 

41. White-crowned sparrow - common 

42. White-throated sparrow - common 
43 o Fox sparrow - uncommon 

44o Lincoln's sparrow - uncommon 

(c) Game Bird Brood Counts 



Species Number of Young Date 

With Female 



Willow ptarmigan 2 July 28 

n 

n 

5 Aug. 1 



n 



3 July 31 

3 Aug. 1 



Spruce grouse 7 or 8 Aug. 4 

" " 2 or 3 Aug. 5 



Sharp-tailed grouse 10 or 11 
" » 3 - 10 



White-winged scoter 1 

Surf scoter 7 

Green-winged teal 4 

American merganser 5 



Aug. 


9 


Aug. 


10 


Aug. 


11 


Aug. 


4 


Aug. 


4 


Aug. 


3 


Aug. 


7 


Aug. 


3 



APPENDIX II 

Annotated List of Mammals Observed 

1. ' Black"! - Ursus americanus - this species was common 

bear , % throughout the study area. 

2. Polar bear - Thalarctos maritimus - photographs were taken 

of a polar bear den in the vicinity of Shagamu Lake. 
Apparently this species frequently comes as far inland 
as Shagamu Lake to have cubs. 

3. Mink - Mustela vison - common throughout the study area. 

4« Otter - Lutra canadensis - common - trails throughout the muskeg. 

5. Red fox - Vulpes fulva - common. 

6. Timber wolf - Canis lupus - uncommon - some scats seen. 

7o Harbor Seal - Phoca vitulina - common along coast - 17 seen 
from Winisk to Shagamu River. 



; 



- 36 - 

Bo Red squirrel - Tamiasciuris hudsonicus - uncommon - apparently 
squirrels had reached a low about three years ago and 
have not been common since - only one individual was 

seen. 

9. Red-backed vole - Clethrionomvs gapperi - common. 

10. Muskrat - Ondatra zibethica - common in larger lakes. 

lie Beaver - Casto r canadensis - common but not as plentiful as it 
had been. The Mishamattawa watershed seemed to have 
a higher density of beaver than the Shagamu although 
availability of food appeared to be equal in the two. 

12 o Varying hare - Lepus americanus - very common . 

13 o Moose - Alces alces - common along river banks and in burned 

areas. 

14<> Woodland caribou - Rangifer tarandus - common throughout study 
area. 

15 o White whale - Delphinapterus leucas - 4 seen along the coast. 

Not as common as it would have been later on in summer, 



LITERATURE CITED 

Ahti, Teuvo.and R. L. Hepburn, 1961. Preliminary Study of Woodland 
Caribou Range in Ontario. Unpublished report. 

Coombs, Donald A. 1954« 

The Physiographic Subdivisions of the Hudson Bay Lowlands 
South of 60° No - Geographical Bulletin #6. 

Hustich, Ilmari. 1957* 

On the Phytogeography of the Subarctic Hudson Bay Lowland. 

- Acta Geographica 16 No. Is 4$ pp 

Simkin, D. W. (1961) - 1961. 

Aerial Survey for Caribou in Hudson Bay Lowland. 6 pp. 

- unpublished. 

Simkin, D. W. I960, (a) 

Woodland Caribou Project, July 1959* Dept. of Lands and Forests 
Fish and Wildlife Management Report #52, June I960 - pp. 1-9. 

Simkin, D. W. I960, (b) 

Aerial Survey for Caribou in Hudson Bay Coastal Area for I960. 
5 pp. unpublished. 





fH 




< 




CQ 




S 




o 




CO 




Q 


■P c 


£> 


O bD 


oc 


H -H 




Ph CO 




a) 3 




HTl O 




Ph CD ,0 




S U -H 




Cti CD Si 




CO > CD flj 




O 4-5 O 




C O -H 




Q) CO X! 




s: as &, co 




O CD 6 CD 




•h k aj ^ 




i-q < o fx, 






- 33 - 
THE WARM WATER SPORTS FISHERY OF 

McGregor bay i960 

by 
H. R. Silva and D. I. Gillespie 

Abstract 

With the collapse of the lake trout fishery in most of 
Georgian Bay and Lake Huron, increased pressure was 
applied to the remaining warm water fishery. The 
McGregor Bay Association expressed concern for the 
maintenance of a sport fishery in the waters of their area«> 
In order to evaluate the situation and with the aid of 
members of the Association, a creel census was conducted* 
The present paper is limited to data collected by Mr. T e 
Buchelt of Birch Island Lodge. During the survey period, 
June 11 to August 31, I960, $,498 fish of seven species 
were caught by 791 fishermen, fishing 5>671 hours. Four 
species, smallmouth bass, yellow perch, rock bass and 
northern pike made up 99«5% of the total creel. A 
seasonal average of 7«2 hours per day were spent angling 
by sportsmen who caught an average of 1.5 fish per rod-hour, 



The creel census, often considered our most fundamental 
method of collecting sport fisheries catch data had its early begin- 
ning in the late 1920* s (Eschmeyer, 1935) <» Today it still ranks as 
one of our most important sources of gathering creel data although 
its use has been curtailed somewhat by the expenses involved. 

It can have qualitative or quantitative value or both, 
depending upon the degree of intensity applied to collecting the 
data. In its earlier form, in the thirties, the creel census served 
primarily as a general record of the fishing, providing the workers 
with merely an inventory of the waters in question. Today, as a 
tool of research and management, it can provide age and species 
composition of the catch, growth rates, reproductive rates, food 
consumption figures and fishing success and effort data. 

A variety of methods have been tried to sample the fisher- 
man's creel varying from the most commonly used printed forms filled 
out by census workers to the use of census stations which rely on 
the voluntary cooperation of the angler. Attempts have been made to 
attain 100% coverage whenever feasible, but in recent years, random 
sampling has provided a more extensive coverage while remaining 
statistically soundc 

Management, today, is finding that the creel census provides 
the data necessary for answering questions raised by sportsmen con- 
cerning the exploitation of a fishery. 

With the collapse of the lake trout fishery in most of 
Georgian Bay and Lake Huron, increased pressure was applied to the 
remaining warm water fishery. Early in 1959 the McGregor Bay Associa- 
tion, an organi.-atio~i consisting of lodge operators and camp owners 
located on that portion of Georgian Bay known as McGregor Bay, 



- 39 - 

expressed concern for the maintenance of a sport's fishery in the 
waters of their area. With little or no information available from 
these waters, we were not in a position to evaluate the situation. 

Members of the Association were visited with the hopes of 
securing their assistance in carrying out a creel census which would 
be prohibitive from an economical and physical standpoint if left to 
the personnel of the District. Several indicated their interest and 
willingness in cooperating with the Department if suitable forms 
were made available and adequate supervision was given. 

Figure I represents the form eventually used in gathering 
the data presented in this report,. This paper, however, does not 
cover the data accumulated by the Association members but is limited 
to the material provided by Mr. T. Buchelt of Birch Island Lodge. 

The Census ed Waters 

The lodge is located at the western end of the area out- 
lined by the McGregor Bay Association and sportsmen using the lodge' s 
facilities fish approximately 30% of this area (map 1). The prox- 
imity of Birch Island Lodge to Highway 68 makes it suitable for spot 
checks and the provision of assistance to Mr. Buchelt in his census- 
ing. 

The waters of McGregor Bay vary in depth from approximately 
75 feet to less than one foot. Topographically the area is character- 
ized by numerous islands differing in size from several hundred acres 
to considerab3.y less than one acre. In addition, numerous shallower 
ridges form an extensive area of shoals. These shoals play an im- 
portant role in the fishery of McGregor Bay as it is over them that 
the major portion of the fishing is carried out. Most of the islands 
and a good portion of the north shore show the typical geological 
transition from the Precambrian granite of the north to the Ordovician 
and Silurian limestones of the south. 

S urvey Meth ods 

Too complicated a census form produces only confusion and 
incorrect data. The form used in this survey (and as mentioned before 
appears as figure 1) , assesses the fishing success for each species 
angled for in the area. This is broken down into two catagories which 
are further broken down into two sub-categories- 

1. resident fishing success 2. non-resident fishing success 
(a) fishing with a guide. (b) fishing without a guide. 

Party size was recorded as was the total number of hours 
spent fishing by each party each day. From these figures, we were 
able to determine the mean (i.e. average) and modal (i.e. mostly 
common occurring) party size and the mean and modal fishing effort 
(measured in hours fishing) . Fishing success was calculated in two 
ways — 

1. fish per fisherman - total fi sh caught per week (formula 1) 

total number of fishermen fishing per week. 



- 40 - 

2. fish per unit effort - total fish caught per week (formula 2 ) 

total number of hours spent fishing per week. 

(Formula 1 is expressed in fish per fisherman while formula 2 is 
expressed in fish per rod-hour). The data were compiled on a weekly- 
basis - the week beginning on a Sunday. 

It may be possible to break down the fishing success into 
species, fishing success as most of the fishermen censused indicated 
that there was very little specific (or selective) fishing taking 
place (i.e. they took what they caught) . However, we have not 
attempted to do so in this report, but may in the future when we 
have succeeding year 9 s data to compare J Finally, weekly variation 
in creel composition and weekly variation in species catch was ex- 
tracted from the data collected. 

The Fishery 

The area under survey for the McGregor Bay Association and 
that portion reported on in this paper is primarily a warm water 
fishery. Smallmouth bass, yellow perch, northern pike, maskinonge, 
yellow pickerel and rock bass make up the more important species of 
the waters. Also caught but less significant in the fisherman 9 s 
creel, are lake herring, pumkinseed, white sucker, white or silver 
bass and bowfin. Part of the McGregor Bay area (North Bay) supports 
one of the Great Lakes last remaining indigenous lake trout popula- 
tion. Few of the anglers from Birch Island utilized this source of 
lake trout and only six (6) appear in this report. However, sports- 
men from other lodges and camps indicated that the lake trout harvest 
of this body of water (North Bay) should be noted as good for the 
I960 season. Relatively few of the trout caught were below five 
pounds in weight and many exceeded fifteen pounds. 

Two thousand eight hundred and sixty-eight smallmouth bass, 
two thousand three hundred and sixty-eight perch, one thousand one 
hundred and ninety northern pike, two thousand and forty- four rock 
bass, twenty yellow pickerel, two maskinonge, and six lake trout make 
up the fishes reported in this creel survey. Approximately one hun- 
dred and fifty of the northern pike exceeded fifteen pounds in weighty 
the maskinonge were 21 and IS pounds, the smallmouth bass were pre- 
dominated by the ten to twelve inch size class and perch and rock 
bass were commonly represented by fish that exceeded eight ounces in 
weight. Figure 2 shows the percent composition of the season's catch. 

Fishing Success and Effort 

The census recorded 791 fishermen fishing a total of 5671 
hours for a mean effort of 7.2 hours per fisherman and a modal effort 
of 6 hours per fisherman. The mean party size was 2.15 men with a 
modal size of 2. Eight thousand four hundred and ninety-eight fish 
of all species were caught for a mean of 1.5 fish per rod-hour. This 
rod-hour success varied on a monthly basis from 1.0 in June to 2.1 in 
July to 1.1 in August. 



- 41 - 

Figure 3 shows the weekly variation in the fish per rod-hour 
or fishing success. Figure 4 shows a break-down of fishing success 
into fishermen with guides and fishermen without guides. Maximum 
success was attained during the week of July 17th, when guided fisher- 
men caught 2.5 fish per rod-hour and non-guided fishermen caught 2.2 
fish per rod-hour. The average for guided and unguided fishermen 
was 2.3 fish per rod-hour. 

Figure 5 shows success measured as the number of fish per 
fisherman. This is broken down into guided and not guided angler 
success. The week of July 10th produced the highest success with a 
combined total of 14<>& fish per fisherman being caught as did the 
week of July 24th. 

Figure 8 shows total weekly rod-hours and mean weekly rod- 
hours. The former indicates that the waters under survey were ex- 
posed to the greatest fishing effort during the week of July 10th 
when 106 fishermen fished 701 hours. If the weeks of June 5th, 
June 12th and August 2$th are excluded as being insufficient samples, 
then the weeks of July 17th, July 24th and August 21st represent the 
periods when the greatest effort per unit fisherman was exerted on 
the fisheries. 

Tables 1 and 2 list the data from which the previously 
mentioned figures were constructed. The presence of a seasonal cycle 
in the availability of lake trout based on yield per unit effort has 
been demonstrated by Fry (1949) « If figure 3 is examined, it suggests 
a similar cycle in a warm water fishery. From a relatively low fish- 
ing success in June, there was a sharp increase to a peak during the 
week of July 17th. A decrease in availability began during the 
following week (July 24th) and proceeded rapidly and uniformly until 
the census was completed. 

Fry (op.cit) points out that the curve produced for lake 
trout is not a typical yearly picture, but might vary from year to 
year. We feel that in the case of the warm water species, a simi- 
lar situation exists and the conditions present during the I960 
season were such that this smooth curve was produced. 

It is interesting to compare this curve with the results 
of a similar creel census carried on in a ecldwater fishery during 
the same period. In this case, the sample size was considerably 
smaller. If a monthly average availability figure is used, we can 
smooth out the cold water fishery curve. Table 3 provides the 
comparison of the two fisheries while figure 9 illustrates the 
difference in availability. Peak availability was during the week 
of July 17th in the warm water fishery while the low point in avail- 
ability for the cold water fishery occurred the following week 
(July 24th) . 

Fry (1939) states this annual cycle in availability of Lake 
Opeongo lake trout is a reflection of the migratory and feeding 
behaviour of the species brought on by the hyrological changes in the 
lake. This appears to be true for our cold water fishery but the 
curve produced in the McGregor Bay fishery does not seem to be cor- 
related to thermal conditions of the water. This may be due to our 
clumping of species to produce an aggregate availability picture. 



- 42 - 

Individual Species 

Four species, smallmouth bass, perch, rock bass and northern 
pike, made up 99«5^ of the season's catch. Figures 6 and 7, a, b, c, 
d, e, show the weekly per cent harvest of individual species and each 
species as a per cent of the weekly catcho 

Smallmouth Bass 

This species made up . 33 "1% of the total harvest. With 
the opening of the bass season on July 1st, this fish quickly assumed 
a leading role in the fishermen's creel. As predicted by South Bay 
Research Station (Anonymous, 1959) > the 1955 year class played an 
important part in this fishery. A large portion of the catch consisted 
of fish measuring ten to twelve incheso With the exception of the 
weeks of July 17th, 24th and 31st, the per cent of the weekly catch 
showed a steady increase from 23 «> 9% during the week of June 26th to 
60.8$ during the week of August 28th. This is shown in figure 10 
where the weekly maximum water temperature has been plotted with 
these figures. 

The week of July 10th produced the greatest harvest with 
528 or 18 . l+% of the bass sample being caught „ 

Perch 

A total of 2368 perch or 27 08$ of the total catch were taken 
during the thirteen week census period. This species was received 
favourably by the fishermen partially due to its large size and 
partially due to the type of fishermen — i.e. non-residents who 
seek it as a game fish. It is encouraging to see more people being 
satisfied with the so called pan fish. As with the smallmouth bass, 
the week of July 10th produced the largest percentage (18.5) of the 
total catch. 

Rock Bass 

This fish, because of its large size, was a favoured 
species among the Birch Island Lodge anglers. Two thousand and 
forty-four or 24.0% of the total catch were harvested. It made up 
the greatest proportion of the weekly catch during the early part 
of the survey, but gradually diminished in importance as the season 
progressed. This change in importance is the reverse to that of the 
smallmouth bass. 

Northern Pike 

Contrary to the 1959 South Bay report (op.cit.), the north- 
ern pike fisheries maintained its former level and only two of some 
1190 caught showed signs of lamprey predation. This species showed 
the most consistant level of % weekly success. The week of July 10th, 
as with the smallmouth bass and perch, produced the heaviest harvest 
of northern pike for the season* 



- 43 - 

Summary and Conclusions 

1. Eight thousand four hundred and ninety-eight fish of seven species 
were recorded during the survey period June 11th to August 31st, I960, 
by 791 fishermen fishing 5>671 hours,, Four species, smallmouth bass, 
yellow perch, rock bass and northern pike made up 99«5% of the total 
creel. A seasonal average of 1*2 hours per day were spent angling 

by sportsmen who average 2.15 men per party and caught an average of 
1.5 fish per rod-hour, 

2. The fish per rod-hour success for June, July and August produced 
a uniform curve which paralleled the cycle of availability for Lake 
Opeongo lake trout and was the inverse of the availability picture 
for a cold water fishery which was censused coincidental in Sudbury 
District o Maximum success, measured as fish per rod-hour, was 
achieved during the week of July 17th when 2.3 fish per rod-hour 
were taken; measured as fish per fisherman the weeks of July 10th 
and July 24th produced 14«$ fish per fisherman. 

3» Smallmouth bass showed an increasing availability throughout the 
census period. 

4« Yellow perch and northern pike along with the smallmouth bass 
were most heavily harvested during the week of July 10th. 

5« The northern pike maintained the most level rate of availability 
of any species. 

6. While lamprey predation in these waters has eliminated all lake 
trout populations other than the one found in North Bay, little 
evidence of predation was found in the species censused in this 
survey. Two large northern pike bore lamprey scars. 

7<> It is difficult to assess a sports fishery on one year's data 
and it is hoped that a similar survey will be carried out over a 
period of years. 

3. We can safely say that the McGregor Bay waters are providing the 
type of angler using Birch Island Lodge's facilities with adequate 
success based on the availability and harvest figures. 

9o Figure 3 indicates that time of year plays an important role in 
this fishery. Optimum success was achieved during mid July in I960. 

10 o Such a survey provides the Department with the facts necessary 
to answer the sportsmen's queries on fishing success and serves as a 
better indicator of exploitation than casual observations do. 

Acknowledgments 

The authors are indebted to Mr. T c Buchelt of Birch Island 
Lodge for his intense interest and cooperation in carrying out this 
creel census. The project was initiated on the recommendation of 
Mr. C. F. Bibby and his criticism of our census form is greatly 
appreciated. We are also indebted to Dr. A. McCombie and J. Budd of 
the Research Branch for their critical appraisal of this report. 



- 44 - 

Literature Cited 

Anonymous, 1959» Angling in South Bay, in 1959? Ontario Department 
of Lands and Forests, Research Branch, Mimeo, 12 pp. 

Eschmeyer, R, W. 1935« Analysis of the Game Fish Catch in Michigan 
Lake, Transc Amer. Fish. Soc. 65 = 207-223. 

Fry, F. S, J., 1949c Statistics of a lake trout fishery, Biometrics 
American Statistical Association, Vol. 5 = No. 1. 

Fry, Fo E. J., 1939. A Comparative Study of Lake Trout Fisheries 

in Algonquin Park, Ontario, University of Toronto Studies 
Biol. Ser„ 46, Pub. Ont. Fish. Res. Lab. 5#» 69 pp« 



- 45 - 



Table Mo, 1 



Weekly Catch By Species 






Date 
Week of 

June 5 
12 
19 
26 

July 3 
10 
17 
24 
31 
7 
14 
21 
23 



Aug, 






June 5 
12 

19 
26 

July 3 
10 
17 
24 
31 

Aug. 7 
14 
21 
28 



June 5 
12 
19 
26 

July 3 
10 
17 
24 
31 

Augc 7 
14 
21 
23 



Species 
Rock Bass 



Number Caught 

7 

84 
111 
107 
300 
394 
269 
246 
275 
103 

83 

49 

16 



Northern Pike 



Perch 



2044 



19 

34 

81 

159 

196 

176 

179 

158 

78 

42 

54 

Ik 



1190 



Per cent of 

Weekly Catch 

100 
81.6 
50 c 2 
27 o 2 

27e3 
25o2 
25.8 
20 o 5 
24.6 
13. 6 
16.3 
13*7 
11.2 



18.4 
15.4 
20. 6 
14.5 
12c6 
I6c9 
14.9 
14.1 
10.3 

8.3 
15.1 

9.8 



Per cent of 
Total Individual 
Species Catch 

c3 

4.1 
5c4 
5.2 

14.7 
19.3 
13 o 2 
12.0 

13 o 5 

5.0 

4.1 

2.4 

.8 



1.6 
2.9 

6.8 
13.4 
16.5 
14.8 
15.0 

13o3 
6.6 
3»5 
4.5 
1.2 



76 


34. 4 


3.2 


108 


27.4 


4<> 6 


271 


24o7 


11.4 


439 


28.1 


18.5 


276 


26.5 


11.7 


390 


32.6 


16.5 


321 


28.7 


13 c 6 


266 


35o2 


11.2 


151 


29o7 


6.4 


45 


12.6 


1.9 


25 


17.5 


1.1 


2368 







- 46 - 



Table No. 1 Cont'd 

















Per cent of 


Date 










Per cent of 


Total Individual 


Week of 


SjD 


ecies 




Number Caught We 


ekly Catch 


Species Catch 


June 5 
12 


Smallmouth 


Bass 










19 
26 








94 


part wko 


23.9 


3.3 


July 3 








361 




32,9 


12.6 


10 








528 




34.8 


18.4 


17 








320 




30.7 


11.2 


24 








383 




32.0 


13.4 


31 








358 




32.0 


12.5 


Aug. 7 








297 




39.3 


10.4 


14 








231 




45o4 


8.1 


21 








209 




58.4 


7o3 


23 








2868 




60.8 


3.0 


Table No 


. 2 


Weekly 


Distribution of An 


glers, Tota 


1 






Rod- 


-Hours, Fish 


Per Angl 


er, Fish Per 




No.' 


Rod-Hour, Mean Rod-Hour : 


er Angler 




Date 


of Anglers 


Rod-Hrso 


Total 


Fish/Angler 


Fish/Rod- Moan Rod- 












Fish 




Hour Hr. per 
Angler 


June 5 




2 




8 


7 


3o5 


88 4.0 


12 




21 




172 


103 


4c9 


.60 8.2 


19 




33 




207 


221 


6.7 


1.1 6.3 


26 




47 




311 


394 


8 4 


1.3 6.6 


July 3 




91 




568 


1097 


10 2 


1.9 6.3 


10 




106 




701 


1561 


14c8 


2.2 6. 6 


17 




79 




448 


1043 


13. 2 


2.3 5»7 


24 




81 




559 


1198 


14.8 


2.1 6.9 


31 




85 




5BS 


1119 


13 2 


1.9 6.9 


Aug. 7 




86 




542 


755 


^cS 


1.4 6.3 


14 




75 




486 


509 


6.8 


1.0 6.5 


21 




60 




416 


358 


6.0 


.86 6.9 


28 




25 




173 


143 


5.7 


.83 6.9 



TOTALS 



791 



5,671 



8,498 



- 47 - 









Table No. 


3 
















A COLD 


WATER 


FISHERY 


A WARM 


WATER FISHERY 




Availabil 


ity 


Availab: 


Llity 


Week of 


Fish/Rod-1 


-Ir. 


Monthly Mean 


Fish/Rod- 


tfr. 


Monthly Mean 


June 5 


.63 








.33 






12 


.5$ 






c49 


,60 




1.0 


19 


.36 








1.1 






26 


.25 








1.3 






July 3 


.51 








1.9 






10 


.32 








2.2 






17 


.32 








2.3 






24 


.16 






.39 


2.1 




2.1 


31 


.47 








1.9 






Aug. 7 


.22 








1.4 






14 


.26 






.26 


1.0 




1.1 


21 


.20 








.86 






23 


o37 
Bcorded in 


Report 




.33 






Species R< 












Sea Lamprey 
Bowfin 

Lake Herring 
Lake Trout 
White Sucker 
Northern Pike 
Maskinonge 
Silver Bass 
Yellow Perch 
Yellow Pickerel 
Sraallmouth Bass 
Rock Bass 
Pumpkinseed 



Petromyzon marinus 
Amia calva 
Leucichthys artedi 
Salvelinus namaycush 
Catostomus commersonni 
Esox lucius 
Esox masquinongy 
Lepibema chrysops 
Perca flavescens 
Stizostedion vitreum 
Micropterus dolomieui 
Ambloplites rupestris 
Lepomis gibbosus 



- 48 - 










v. 



< 



a. 

c 



Pi 




* \ i , 



<? 



hJ^) x 



< 

UJ 

< 

Q 
Ul 

u 

> 
cr 

CO 



> 



,v" 



c\ 



~) 



- 49 - 

McGregor bay association 
daily creel census 



INDIVIDUAL 




NOo OF FISH REPORTED 




_ 
EFFORT 


REMARKS 
OR 






-p 




p I-p 


p 




P 


i -I- 5 






WEATHER 




p 


C -P 




G 




G 




g 




G 




C 




CO 






G 


CD 3 


CO 


CD CO 




CD 




CD 


CO 


CD CO 


P 


CD P 




c£ 


CON- 




O 


T3 o 


CO 


Xi CO 




Xi 




-a 


CO 


T3 CO 


3 


T3 2 


<3 




-P u 


•H 1h 


P ctf 


•H ctf 


p 


■H 


p 


•H 


P ctf 


•H Ctf 


P o 


•H O 


cti 


i-1 


DITIONS 




G E-i 


CO Eh 


d CD 


co m 


c 


CO 


c 


CO 


C CQ 


CO PQ 


G G 


CO U 


& 


o 




CD 


CD 


CD 


CD 


CD & 


CD x: 


CD 


CD 


CD 


CD 


CD E-t 


CD Eh 


o 


s 






XS CD 


f£ CD 


xi • 


Cti • 


X) o 


of, o 


Xi CD 


0-4 CD 


X} X 


0^ ^ 


xs 


CnJ 


ps 


.«: 






•H JsJ 


1 Xi 


■rig 


i S 


•H *-, 


1 ^ 


•H ^ 


1 ^ 


■H O 


1 o 


•H » 


1 • 


\ 


— 






CO CT5 


G ctf 


CO • 


a • 


CO CD 


C CD 


CO -H 


a «H 


CO O 


G O 


CO ft 


G G< 


• 


• 






CD h-1 


Okh 


CD CO 


o co 


CD Ph 


o a. 


CD Ph 


oa, 


CD &4 


O 0^ 


CD CO 


O CO 


o 


o 




DATE: 


cd 


s 


Pi 


3 


Cr4 


a 


pj 


s 


0^ 


S 


0", 


3 


s 


s 






GUIDE 


































NO 


































GUIDE 


































GUIDE 




































Figur 


e 1. 


mm — 


Form 


1 

used 


foi 


- McC 


1 

iregor Bay creel 


census. 







40,- 



30 



20 



10 



o 

-p 

ctf. 

o 

rH 
ctf 
P 
O 

Eh 

P 

CD 
O 

U 
CD 

P-. 



O o O o 

2363 




6 



20 



2 



It 



Small- 
mouth 
Bass 



Perch 



Rock 
Bass 



Pike Lake Walleye Maskinonge 
Trout 



Figure 2» — Species as per cent of Total Catch 



- 50- x 
x/ 

/ 

/ x N 

/ \ 

/ \ 

/ \ 

/x ^ 

/ \ 

/ \ 

/ \ 

/ \ 

/ \ 

/ \ 

/ \ 



/ 

/ 

/ 



/ 



/ 



/ 
/ 

/ 



\ 
\ 



\x 
\ 



/ \ 

/ x \ 

/ \ 

/ V 



\ 



x \ 

/ x X 

/ x \ 

/ \ 

/ V 



\ 



i 1 j 1 j i i I » i I I r~ 

5 12 19 26 3 10 17 24 31 7 14 21 2B 
June July August 

Figure 3„ -- Temporal Fishing success (Fish/Rod-hour) all Fishermen, 



2.5 



2.0 



o 

ac 

i 

-d 
o 

u 
o 

CO 
•H 



1.5 



1.0 



- 51 - 



* o 
o X ' 



X 



/ \ 



/ 



X 



/ 



'/ 

/ 

/ 

/ 



X 



o/ 



/ 

x /° 

/ • 
• / • • 

/ 
/ 



s 



•x 



\ X o 

\ 
\ 
• \ 



\ 



\ 



\° 

w 

I, 



X 



\ 



\ o X 

\ • 
\ . 

\ ■ 

\ e 

\' 

\ 



X 

k 

\ 
*\ 



\ 



\ 



X 



.X 



oX 



\ 



■• Non-Guided Fisher- 
men 

Guided Fishermen 



5 12 19 25 3 10 17 24 31 7 14 21 
June July August 

Figure 4. — Temporal fishing success (fish/rod-hour) guided vs 

non-guided fishermen. 



23 



20 


























- 52 


— 










— 
































LI Guided Fishing 


18 


*- 






Non-Guided Fishi 

X 








— ! 








16 




1 . 












14 






, 






X 














X 








X 


















X 




X 






X 










12 






X 
X 




X 

X 

X 




X 
X 
X 






X 
X 
X 






X 

X 
X 




10 






X 




X 
X 




X 
X 






X 
X 






X 
X 










X 










































X 




X 






X 






X 


















X 






























8 






X 




X 




X 
X 




X 

X 






X 
X 






X 
X 






X 
X 


















X 




X 




X 




X 






X 






X 






X 




— 1 








£ 




X 




X 






X 




X 




X 




X 






X 






X 






X 




X 












X 




X 






X 




X 




X 




X 






X 






X 






X 




X 
















X 




X 






X 




X 




X 




X 






X 






X 






X 




X 










X 


4 






X 




X 






X 




X 




X 




X 






X 






X 






X 




X 










X 








X 




X 






X 




X 




X 




X 






X 






X 






X 




X 






X 






X 










X 




X 






X 




X 




X 




X 






X 






X 






X 




X 






X 






X 


2 








X 
X 

X 




X 
X 
X 






X 

X 

x 




X 
X 
X 




X 
X 
X 




X 

X 
X 




I . 


X 
X 
X 






X 
X 
X 






X 
X 
X 




X 
X 
X 






X 
X 
X 






X 
X 
X 



12 19 26 10 17 
June July 

Figure 5. — Temporal distribution 



24 31 7 14 21 28 

August 

of fish per angler. 



- 53 - 



SO 

$ of 
Total 
S.M, Bass 
Catch 50 



Figure 6»A - Smallmouth Bass 



40 



20 



100 



30 



Figure 7 .A 



% of 
Weekly- 
Catch 



60 



40 



20 



5 12 19 26 3 10 17 24 31 7 14 21 23 
June July August 






- 54 - 



SO 



Figure 6.B - Rock Bass 



60 



% of 

Total 

Rock 

Bass 

Catch 



40 



20 



100 



B0 



60 



% of 

Weekly 

Catch 



40 



20 



Figure 7»B 



5 12 19 26 3 10 17 24 31 7 14 21 2$ 
June July August 






BO 



% of 
Total 
Perch 
Catch 



60 



40 



20 



100 



BO 



60 



% of 
Weekly- 
Catch 



- 55 - 



Figure 6.C - Perch 



rrr 



Figure 7-C 









40 



20 



5 12 19 26 3 10 17 24 31 7 14 21 28 
June July August 



80 



% of 
Total 
Pike 
Catch 



60 



40 



20 



100 



80 



% of 
Weekly- 
Catch 



60 



40 



20 



- 56 - 



Figure 6oD - Northern Pike 



zn 



Figure 7cD 



m 




























































































5 1 
Ji 


2 1 

me 


9 2 


6 


3 


1 
Ju 


1 

iy 


7 2 


!4 


3 


1 


7 


1 
Augu 


4 

St 


2 


1 


28 



so 



- 57 - 

Figure 6»E - Yellow Pickerel 



60 
% of 
Total 
Yellow 
Pickerel 
Catch 

40 



20 



100 



80 






1 

■ ■ ■-■ 

I 



Figure 7«E 



% of 
Weekly- 
Catch 



60 



40 



20 






14 21 2"8 



19 26 3 10 17 24 31 7 
June July August 






- 58 



* 



x 
'\ 

/ \ 

I \ 

I \ 

I \ 



X 



..Total Weekly Hrs. 
jcMean Weekly Hrs. 



mo 



6- 

Hrs. 

5 



4 



/ 
k x 




700 



600 



Hours 



500 



400 



300 



200 



100 



5 "12 19 

June 
Figure 8„ 



July 



31 ' 7 " 14 ' 

August 



23 



Temporal fishing pressure in hours — mean and 
total hours o 



- 59 



co 


CO 





CD 


•H 


•H 


O 


O 


CD 


CD 


a 


A 


CO 


CO 


Sw 


U 


<D 


CD 


-P 


4-5 


cC 


Cd 


13 


^ 


S 


T3 


u 


H 


cd 


O 


3 


O 



X 



X 



\ 
\ 



s 



>!- 




vD 

rH 
O 

CV 

UA 
CV 

-co 



H 
-± 

to 

CV 

H 

cv 
~-^ 

H 

o- 

H 

cv 

o 

H 



3 



CV 
ON 



CV 

rH 



U"\ 



ON 

CV 

"cv 

CV 

H 

to 



H 



CD 
-P 

cd 





O 




o 


!h 


cd 


CD 




,£) 


C 


£ 


•H 


CD 





P 


co^ 


ft 


Q) U 


CD 


•H 3 


CO 


O O 




CD JZ 




Ph 1 




to T3 




O 




X ^ 


-P 


C0\ 


CO 


•H ^ 


2 


^H CO 


bO 


•H 


3 


<H <H 


<£ 


O--" 




>• >, 




P ^ 




•H CD 




H X 




•H CO 




,£! -H 




cd H 




rH 




•H U 


!>> 


a cd 


H 


> p 


3 


a cd 


>s 


£ 




CD 




x s 




p s 




cd 




«H % 




o 




cd 


CD 


pi 


C 


O T3 


3 


co a 


•"3 


•h cd 




u 




cd >> 




Ph rH 




S CD 




OX 




O CO 




•H 




< CH 




1 

1 




O 


!>» 




cfl 


CD 


S 


rH 




3 




hO 




•H 




Pm 



o 

o 



O 



CV 



U 
o 

EC 

co u -d 

•H CD O 

Cn Qh Cd 



o 

o 
rH 



- 60 - 



66 



_ # Maximum Water Temperature °F 

- Weekly Bass Harvest % of Total Catch 



65 



64 

Max. 
Temp. 

Op 

6 3 60 
Bass 
as 
% 
of 

62 50 

Weekly 
Catch 



61 40 



60 



30 




x 
/ 



/ 



X 

/ 



/ 



/ 



X 

/ 



x / 



-x- 



59 20 



10 



June 



15 *26 '3 '10 '17 '24 "31 '7 ' H r 21 ' 28 
July August 

Figure 10. — Smallmouth Bass Harvest related to maximum water 
temperature.