Report on the Animals Within the Taramakau Catchment

REPORT ON THE ANIMALS WITHIN THE TARAMAKAU CATCHMENT

L.W. Best and P.S. Crosier

Protection Forestry Branch Report No. 74 Copy No.

Date: February 1970

Protection Forestry Branch,

Forest & Range Experiment Station, RANGIORA.

s ) 1.

REPORT ON THE ANIMALS WITHIN THE

TARAMAKAU CATCHMENT

FOREWORD

Authorship of this report warrants a brief comment.

Technician L.W. Best managed the field survey, made the first analysis of data, and wrote the first draft of a report. He then resigned to take a position with C.S.I.R.O. in Australia. At this time it was decided to include parts of reports written by Scientists

C.N. Challies, on condition of deer, and I.L. James (F. & R.E.S.,

Napier), on correlation of opossums with damage in forest canopies, because these were very pertinent to the objects of the survey. It thereby became necessary to assign a new face to the tasks of writing up, collating the different MSS, and checking all the records and computations. This was undertaken by Philip S. Crosier, a Zoology student at University of Canterbury, who had been employed on the survey, during November-December 1969. In addition, the contributions of

R.E. Lambert, J.T. Martin, M.J. Barnett and I.G. Douglas of the

Animal Research Section warrant a special word of appreciation, as does the co-operation, advice and helpful criticism from J.A. Wardle,

J. Hayward and G.R. Evans, who managed the vegetation surveys.

C.L. Batcheler

Head, Animal Research Section,

Rangiora. 2.

SUMMARY

1. A survey was undertaken in the Taramakau catchment to determine the status of red deer (Cervus elaphus), chamois (Rupicapra rupicapra), opossum (Trichosurus vulpecula), and hares (Lepus europeaus) in

Protection Forestry planning.

2. The survey involved running random pellet lines in grassland and forest, along with shooting and autopsy of red deer.

3. The highest red deer pellet density was in the forest of the

Deception area. The Top Taramakau forest also had a high pellet density. The True Right grassland had the highest grassland pellet density.

4. Chamois pellet densities were highest along the Aickens and

Barron ranges.

5. Opossum pellet densities were highest along the True Right of the Taramakau river and in the Deception river.

6. Hare pellet densities were highest in the Top Taramakau grass land, Goat Hill and along the Kelly range.

7. The areas with the highest deer pellet densities were highly associated with the critically damaged or highly modified vegetation associations, viz: short-scrub hardwood, kamahi forests, and the

Chionochloa pallens - f · austral is grasslands.

8. There is a highly significant correlation between opossum pellet frequencies in plots and the percent browse damage on canopy trees in the rata-totara and kamahi forests.

9. High hare pellet frequencies were highly associated with grass land areas containing Chionochloa australis, C. pallens and Celmisia

species.

10. The Taramakau red deer population is at a density where heavy

pressure has been exerted on preferred food resources and it is now

utilising non-preferred food resources.

BARRO BARRO

\. \.

j' j'

\ \

/ /

r~ r~

I I

( (

'"'~ANGE '"'~ANGE

/ / I I

\ \

/ /

\ \

__ __

·~-

KELLY KELLY

/~ /~

/~· /~·

\ \

£? £?

\ \

KELLY KELLY

) )

.. ..

I I

/ /

Figure Figure

"" ""

.. ..

BALD BALD

''·-···------. ''·-···------.

/.. /..

RAN~_....,....-

·.___. ·.___.

31 31

'1(,._:_:_::;F~N '1(,._:_:_::;F~N

~ ~

'-... '-...

.. ..

--~··-·· --~··-··

) )

RANGE RANGE

. .

\ \

------: ------:

1. 1.

\ \

----~-· ----~-·

\ \

-----~. -----~.

Map Map

\ \

----.._/ ----.._/

: :

\22 \22

11 11

~---·-·x----ALEXANDER ~---·-·x----ALEXANDER

DECEPTION DECEPTION

...-----

of of

----

~ ~

·--

~ ~ 18 18

survey survey

~ ~

.. ..

/' /'

I I

MT MT

___/( ___/(

_,,...--

// //

RIVER RIVER

ALEXANDER-··------

_/--· _/--·

AICKENS\_RANGE AICKENS\_RANGE

area, area, I I

\_ \_

---

\. \.

/.. /..

\ \

I I

-1-----·---

-----

I I

\ \

I I

\ \

the the

I I

\~ \~

j j

\ \

34 34 \} \}

] ]

/ /

16 16

RANGE RANGE

j j

Upper Upper

I I

....______) ....______)

__..,,.-· __..,,.-·

Taramakau Taramakau

./ ./

_,j _,j

21;r;· 21;r;·

__ __

---·· ---··

I I

__..-_.--

· ·

I I

'\ '\

.· .·

/ / -+----7---

Catchment Catchment

\ \

// //

- \

I I

/MAIN /MAIN

., .,

. .

I I

/ /

/""/ /""/

2 2

• •

~~ ~~

MICHAE MICHAE

DIVIDE DIVIDE

',20 ',20

./' ./'

......

,/ ,/

CREEK CREEK

--·-··-----...__ --·-··-----...__

./ ./

.. ..

/ /

.. ..

5 5 6

/ /

.. ..

• •

t t

/ /

N N

line line

.////~./ .////~./

tronsec tronsec

/ /

, ,

t t

~ ~

------

-~-----

30 30

/ /

-··----.... -··----....

, ,

/ /

! !

: :

I I

24 24

/' /'

'"' '"'

/ /

\_; \_;

-~~ -~~

( (

: :

HARPERS HARPERS

I I

, , I I 11. It is recommended that the deer control programme be intensi

fied in the Deception forest area, the Top Taramakau area and to a lesser extent, the Otehake river headwaters. The use of helicopters and 1080 poison on broadleaf is suggested as the method of control.

12. Chamois could probably be controlled more economically with

the use of helicopters by venison recovery teams.

13. Opossum control needs more emphasis than it has had in the

past. Three proposals are put forward: the exclusion of private fur

collectors from the critical areas where no interference in the control

programme can occur; an intensive winter 1080 poisoning programme; and

th~ integration of the poisoning programme with current opossum research

being done at this Station.

INTRODUCTION

Approximately 100,000 acres of the headwaters of the Taramakau

catchment (42°50 1 s - 171°41 1 E) were surveyed to estimate the status of

red deer (Cervus elaphus), chamois (Rupicapra rupicapra), opossum

(Trichosurus vulpecula) and hare (Lepus europeaus) in Protection Forestry

planning.

The southern boundaries of the area were Harpers Pass and the

Main Divide south-west to Mount Rolleston, north along the Barron and

Kelly ranges, and west along the Bald range to opposite Hut creek on

Mount Alexander. The main ridge of the Alexander range from Mount

Alexander east to Harpers Pass formed the northern boundary. Most of

the area south of the Taramakau river is part of the Arthurs Pass

National Park (Figure 1).

The survey began on 7 January 1969 and the work undertaken by the

Animal Section was finished on 21 February 1969. A total of 220 man-

days were spent on survey; 60 man-days were occupied with shooting

and autopsy of red deer, the other 160 man-days were spent running 4.

34 pellet frequency lines.

HISTORY OF THE ANIMALS WITHIN THE CATCHMENT

1. Red Deer

The first red deer in the Taramakau catchment were probably pro geny of animals liberated in the Poulter valley in 1908 and 1909.

Descendants of deer liberated in the Rakaia (1903, 1906), Mount Tuhua

(1903, 1906), Lake Kaniere (1906), the Grey catchment (1898) and the

Nelson district (1851 to 1922) may also be in the area. The liberation dates and known intermingling of the herds (Forbes, 1924; Harris and Logan, 1967) suggest that red deer were established in the survey area by 1925-30.

In 1930 the Government removed protection from red deer and in

1931 the Department of Internal Affairs began culling operations.

With the exception of the World War II period, shooting operations have continued up to the present day. Appendix 1 lists the kill figures obtained by Government shooters during the post-war control operations. The survey area lies within the Taramakau shooting block.

Red deer in the catchment have been and still are being hunted by sportsmen and commercial venison recovery firms (Appendix 2).

Most of the regular hunting under the auspices of the Arthurs Pass

National Park has been on the Kelly and Barron ranges (Appendix 2a, b).

Commercial recovery of venison accounts for the greatest number of deer killed outside the Government control programmes (Appendix

2c, d). Most of these animals were taken by helicopter from the alpine grasslands of the Alexander range.

2. Opossum

A list of liberations of opossums in the Taramakau and surrounding

area is given in Table 1 (Pracy, 1962). Pracy carried out a survey

of the opossum population in the Taramakau catchment in 1949 and 1950. He found that opossums were well established below Inchbonnie on the true right of the river but were absent up river beyond this point.

They were also present in the head of the Taramakau and had probably entered the valley over Harpers Pass from the Hurunui river in Canter- bury. On the true left of the Taramakau river opossums were well established at Townsend creek and were present down to and up the

Otira river to about the mouth of the Deception river. These animals were probably derived from an unrecorded liberation at Townsend creek.

TABLE 1 - Liberations of Opossums around the

Taramakau Area (from Pracy, 1962)

Year Colour Locality

1891 Black Hurunui river 1920 Black Haupiri river 1922 Black Poerua river 1923 Black Griffin creek 1926 Black Otira ? Mixed Upper Taramakau ? Black Bald range-Taramakau ? Black Ahaura river

The true left of the Otira was colonised from opossums liberated

at Kelly creek in 1926 and also from animals crossing the Taramakau

river on the railway bridge below Jacksons.

The distribution of opossums on both sides of the Taramakau river

at Townsend creek was checked by Logan and Mangos (1953). Their

results agreed with the distribution found by Pracy.

Opossum protection was removed in 1947. Since then, opossums

have been trapped or poisoned for furs and control purposes. From

1951 to 1961 a bounty of 2/6 (25 cents) an animal was paid to encourage

control. Control programmes in the Taramakau were run in 1964 when 6. the Kelly range was poisoned and in 1968 when an experimental poisoning was undertaken from Taverners Flat to Michael creek. Taking opossums

for furs has probably fluctuated in intensity with the demand for

pelts. There has been intensive trapping and poisoning in the area

in the last three to four years.

3. Chamois

Chamois were liberated in the Mount Cook area in 1907 and 1914

and were well established by 1920 (Thomson, 1922). Chamois probably

arrived in the Taramakau area about 1925. This date is estimated

from the dispersal rates given by Caughley (1963) and Christie (1963).

Protection of chamois was removed by the Government in 1930 and

in 1936 the Department of Internal Affairs began control operations.

The majority of chamois killed in the Taramakau have been shot by

Government shooters (Appendices 1 and 2). Records show that private

hunters in the National Park have not killed many animals (Appendix 2a,

b).

4. Hares

Hares were liberated on Banks Peninsula, Canterbury in 1851

(Donne, 1924), but many unrecorded liberations have also been made.

By the early 1920's hares were common throughout the South Island

(Thomson, 1922). Although their time of arrival in the Taramakau

catchment is not known they may have been the first of the above intro

duced animals in the area.

METHODS

(i) General

Data on animal use within the areawere collected by recording the

faecal pellet frequencies in mil-acre plots along 34 transect lines.

Faecal pellets will be referred to as pellets or pellet groups. The

transect lines ran from a creek or river bed to the open mountain tops. Each line was initiated from a point chosen from a map by restricted

0 randomisation. The starting points and lines running at 5 from the

slope, were recorded on aerici.l photographs for location in the field.

Transect lines were divided into either forest or grassland. The

forest lines ran to a limit where there was a mixture of 50 50 alpine

scrub and grassland, or if grassland was absent, to the top of a main

ridge. The grassland lines were initiated from the end of the forest

lines, and ended where the vegetation cover was absent, or where a

main ridge was reached.

The transect line was followed along a compass bearing using a

60 foot running line. Every 60 feet the presence or absence of pellets

of each species in a mil-acre plot (3' 80511 radius) was recorded. The

minimum unit of faeces was defined as a single pellet with a complete

outer case.

(ii) Forest lines

1. The distance to the nearest group of deer pellets was recorded

on the 5th plot and every 10th plot thereafter (5th, 15th, 25th, ••• ).

A group was defined as six or more pellets with a complete outer case,

belonging to one defaecation.

2. Pellets were searched for in a circle up to a radius of limit

10 feet.

3. A forest vegetation survey was undertaken by a vegetation

reconnaissance party on every 2nd and 9th plot thereafter (i.e. 2nd,

11th, 20th, ••• ).

(iii) Grassland lines

1. Vegetation cover was recorded on the 5th and every subsequent

10th plot. This work was carried out independently of the alpine

grassland party.

2. A tape was run at right angles to the transect line, and 25

points 6 inches apart, recorded on either side of the line. 8.

3. The species of plants and their relative heights were recorded in 1 foot classes lying directly below the 50 points. Each point was selected using a plum-bob.

4. To minimise disturbance of the vegetation the first point recorded was taken two feet either side of the transect line. The altitude, aspect and slope were recorded.

The vegetation height classes were:

Class Height

0 0 - 6 11

11 1 6 11 - 1I6

1 11 1 11 2 1 6 - 2 6 3 2 1 6 11

(iv) Analysis of data

(a) Distribution of animals

The distribution of animals throughout the catchment was established

from the distribution of frequency and/or density of their pellet groups.

This method gives a relative index of density, it is not an absolute

measure of animal density. Providing the measurements are taken in

the same season of one or different years, comparisons of distribution

can be made between areas. Pellet decay rates are assumed equal when

comparing two different areas within the same survey area. The infer-

ence of pellet density is therefore that when the density is high the

animal use * of the association is high, and vice versao

Use is referred to as the relative amount of time spent in a unit area of a vegetation association. In grasslands the use is indicative, on evidence from Cupola Basin, of the relative feeding pressure/unit area, while in forests a greater proportion of time may be spent on non-feeding activities, although pellets are present. (1) Pellet frequencies: The pellet frequency for each forest and grassland section of the lines was determined from the number of plots with pellets/number of plots in the line.

(2) Pellet densities: The density of red deer pellet groups per acre was calculated from the distance measurements to the nearest group of pellets from the plot centre: 2 2 2 4356on; (r + • • • + r + (N - n)R ) groups/acre Density (D) = 1( 1 n where,

n = number of N point distances which are equal to or less

than R.

N = total number of sample points.

+ r are the n distances. n R = radius which includes a group at half the number of sample

points

(Batcheler, pers. comm.)

The pellet frequencies recorded in the whole survey area or any

part of it were transformed into a density from:

Local density (df) = D (log( 1 - f)/log( 1 - F)) groups per acre

where F is the frequency for the whole survey area, and f is

frequency for any part of it.

For chamois, opossums and hares relative densities were computed

from the presence or absence of pellets in mil-acre plots from:

Density ( d) = -log ( 1 - f) e (b) Animal - vegetation interaction.s

The pellet frequency for each vegetation plot was determined by

lumping 10 mil-acre plots, 5 plots from below including the vegetation

plot, and 5 plots above. Analysis of the forest plots involved using

one plot twice because of the 9 plot spacing of the vegetation plots.

The 5th mil-acre plot above the vegetation plot became the first mil-

acre plot of the next series. 10.

The vegetation plots were divided into 12 different groups; 7 were forest groups and 5 were grassland. The names of the vegetation groups are taken from the association analyses of Evans (1969) and

Wardle and Hayward (1969) (see Appendix 3). Differences in the naming of the associations are due to different recording techniques in the field.

Deer and opossum interaction with the forest associations were

tested by using a 2 x 7 contingency table. Deer and hare interaction

with the grassland and subalpine scrub associations were tested using a

2 x 6 contingency table. The contingency tables give a simultaneous

two-way test of the observed relative use by a species of a given

association, compared with the expected use by that species of all

associations and all other species of that association (modified from

Batcheler, 1961). Any vegetation association preference the animals

may have will be shown by the abundance of pellets being greater than

the expected value.

Arbitrary levels of relatively high, medium and low pellet frequency

were chosen to try and give an indication of the vegetation association

that species may prefer. The levels were defined as:-

(i) High use of an association by a species if pellets occurred

in 20% or more of the plots than the expected value.

(ii) Medium use of an association by a species ~f pellets occurred

within approximately 20% of the expected value.

(iii) Low use of an association by a species if pellets occurred

in 20% or less of the plots than the expected.

Opossum browsing was estimated on all canopy species in four

classes (heavy, medium, light and no browse) at more than 1,000 observation

points placed regularly along the transect lines (James, 1969). All

classes of browsing visible on each canopy species were recorded;

that is, if at an observation point some trees of a species were heavily 11. browsed and others lightly browsed, both heavy and light browsing were recorded for that species. The data were summarised for each transect and for the 7 forest communities as an index which varies from 0 to

100% browsing: + M + L 100 Hn n n66 33 Browsing index = N. H is the number of trees heavily browsed, M is the number where n n medium browsed, L , the number lightly browsed, and N the total number n of observations.

Three hundr~d and thirty-six red deer were shot and autopsied in the survey area to determine whether the condition of the individual animals is related to depleted or abundant food resources (from Challies, 1969).

Two measures of condition, length of lower jaw and kidney fat index were noted along with total body length. Jaw length indicates

I total body size and may sum up the animals plane of nutrition during its life. Only yearlings and 2 year old animals are considered in the report. Kidney fat index (Riney, 1955) was used to estimate the condition of the animals (assuming it to be synonomous with fatness).

DISTRIBUTION OF THE ANIMALS

1. Red Deer

The catchment was divided into 5 sections to facilitate comparisons

between different areas (Table 2).

The forest in the Deception area had the highest pellet density

for the catchment (Table 3). The Top Taramakau forest also contained

a high pellet density. Low densities were recorded in the Otehake and

Otira forest areas. In the grassland the True Right area had the

highest pellet density. The next highest density occurred in the

Otehake while the Top Taramakau, Deception and Otira areas had compara- Figure

25L

)tl~~L

Frequency

100

and

~l-

distribution

_//

·=

Red

1Jli

deer

~2Yf57

pellets

'T-168-023

-f

YL;.~

2nl:l

V////J

f .,,Q.

lL l5i '.:)

>- 0

~ w

t::: ~

.J 0

RED

IOOlai

..,,

U> 0

-I

O•

r U>

(j) I

DEER

=Sample

U0<

""mbO'

miles

size

. 12. tively low densities. The Otira area had the lowest density in both the forest and the grassland sections.

For the entire catchment the average pellet density was 120 pellet groups per acre. The density is similar to the value found by Lambert and Batcheler (1969) for the Griffin Creek/Rocky Creek area of the

Taipo river. Figure 2 indicates graphically the pellet frequencies for the forest and grassland sections of each transect line. There appears to be a tendency for higher pellet frequencies to occur in the heads of valleys, and in the forest areas, compared to the grassland areas.

TABLE 2 - Giving Area Boundaries and Line Numbers

Area Limits of Area Line Nos.

True Right Hut creek east to 2, 4, 7, Michael creek 10, 17, 25, 29

Top Taramakau Both sides of Taramakau 1' 9' 20' from Michael Ck and the 24, 28, 30 Otehake to Harpers Pass

Otehake Otehake river area and 5, 15, 16, 19, Pfiefer creek 21, 23, 32t 33, 34

Deception Deception river 8, 18, 26

Otira True left of Taramakau 6, 11, 12, 13, west of Otira mouth to 14, 22, 27, 31, opposite Hut Ck, includ 35 ing Otira river 13.

TABLE 3 - Showins Occurrence of Red Deer Pellets and Estimates of the Absolute Densiti 9f fellet Groups Eer acre - Taramakau Catchment

Area Forest Grassland Combined No. of Plots Freq Dens Freq Dens Freq Dens Forest Grass- land True Right 25.2 156 22.0 141 24.4 159 330 100 Top Taramakau 34.o 236 15.1 93 26.9 178 376 225 Otehake 12.1 73 . 18.5 113 13.3 80 609 140 Deception 36.4 254 15.0 93 30.0 203 187 80 Otira 10.5 63 11.1 67 10.6 63 555 162 Average 19.9 127 15.8 99 18.9 120 Total 2057 707

TABLE 4 - Showing_ Occurrence of Chamois Pellets and Estimates of the Absolute Densit1 of Pellet GrouEs per acre

Area Forest Grassland No. of Plots Freq. Dens. Freq. Dens. Forest Grass- land

True Right o.6 * 3.0 30 330 100 Top Taramakau 0.26 * 0 0 376 225 Otehake 0.98 10 2 .1 21 609 140 Deception 0 0 10a0 105 187 80 Otira 0.7 * 16.6 181 555 162 Average 0.5 2 6.3 67 Total 2057 707 * The density value, transformed from the % frequency, was too low to be of any significance

2. Chamois

The distribution of chamois pellets was associated predominantly

with the alpine grassland, and to a lesser extent with the subalpine

scrub (which is included in the definition of forest). Small numbers

of chamois - in groups of six or less - and a low density of chamois

pellets, were observed over most of the grassland and subalpine scrub 84

Figure

'----=-r

Freque;Sa~tribution

"-----

rc-1~8L5

,,----.I

-~~-

26>·~,

:/V\(~\

)/~

_,,

t"-,,_::

6-/

30~

f6{21)

Jl:_

Cham1±=

33[

/'l'

,00

"--623

p~lets

69.

~8365

::l

r a:: UJ u..

:::> w

w u >

i;J 0

,00

CHAMOIS

~L.

5 6

-I

(fl

..,,

r40l

s;: o z

G'I

)> (fl (fl

~Sample

''"'"='"

-24~

2miles

size 4.

\~ ,r4730L7209lcL . .. / 1·0 ~ \ \ ~100 \30 . 2 . ~ 11- \_____.A u.; 65 ,,,-1 ==------r

~l I

59 30 f ; 6_/L-_ ~ ~I J

003 ~~~~f' ~ o '~19I ~"'------

2 [D_~~b 32~ 20 OPOSSUM

??- >- 0 u 100

(5::> 75 t45···30 =Samplesize ~,0 l 23~ 0 ~50 r- 25 I· 1.1.l . _J 0 ... uJ .,, G': 0 20 0.. 0 JJ :{ JJ )> rn U1 (f> (f> -l r N )> . ~ ~2~rs1 60 5 \ , 16 ,, 20 I c.- f 5 6 - '"'"""'" 0 I 2 rn:les

---~ / 14.

areas of the catchment. The highest density of chamois pellets in

the forest occurred in the Otehake (Table 4). In the grassland the

highest pellet densities were in the Deception on the Aickens Ranges,

and in the Otira along the Barron Range (Figure 3).

3· Opossums

Opossums were widely distributed within the forested areas of the (_F,'J·~) catchmen~. Some pellets did occur in the Otira grassland section,

though they were of no real significance. The True Right and Deception

areas held the highest densities (Table 5). A high pellet density

was recorded in the head of Pfeifer creek. The Top Taramakau, Otehake

and Otira areas had lower densities. On the True Right of the Taramakau

river the highest centre of population density was in the vicinity of

Hut creek.

TABLE 5 - Showing Occurrence of 0Eossum Pellets and Estimates of the Absolute Density of Pellets per acre

Area Forest Grassland No. of Elots Freq. Dens. Freq. Dens. Forest Grass- ' land

True Right 35.4 437 0 0 330 100 Top Taramakau 19.4 216 0 0 376 225 Otehake 19.7 240 0.14 * 609 140 Deception 40.6 521 0 0 187 80 Otira 21.1 237 2.5 25 555 162 Average 27.2 330 '0.53 5 Total 2057 707 * The density value, transformed from the 7'; frequency, was too low to be of any significance. Figure

27L~o

)

Frequency

vl4

1' ~

/;351

22 11~f~fl3~s~

100

LLi

and

r77"7'

1 L I/

LV :,:

.. 100

~to 47

distribution

------2J5L_

30/

JYBl t

•

·n

20 0\

rro

}

Hare

~~2yt7 JL

15 '2) 75

pellets

o /

2 9 \

/'")

LJ t

l ~

100

"~023

roe

'--'20 4ul_~l----

" 32~

l~LD)•

--1'.LaL

2bf,,

__/~';g'

1C?Ql-~

6 ~IOOlni I- LI.. li! w w

a. w .J

r f

HARE J/1

.•

. '·

,, l.fl rn 0

:0 -I

5..

;

:0 G>

l.fl t.r>

i;;:: )> o z

•

Uo.°"mb~

Sample

miles

size

20 _

--~ 15.

4. Hares

The highest frequencies of hare pellets were found in the grass land of the true left of the Taramakau, the Kelly range and Goat Hill

(Figure 5). The grasslands of the true right of the Taramakau,

Deception and Otehake Rivers had very low or zero hare pellet frequencies. Pellets were recorded in the forest associations on the edge of the river flats, especially the flats of the Taramakau river from the Otira river to Locke Stream, and on the Otira river flats.

5. Domestic Stock

About 300 head of sheep and six to ten head of cattle were grazed on the Taramakau river flats. The cattle utilised the area from the

Taramakau-Otira confluence to the Pfeifer creek outlet, while the sheep used this area as well as the flats up to Locke Stream. The presence of sheep was indicated in some of the forested river flat areas.

The grass flats at the Deception River mouth held about 100 head of sheep and 20 head of cattle. These animals did not go up beyond the lower river flats.

The river flats below the Otira township and on the true left of the Taramakau river below the mouth of the Otira were fenced and farmed with sheep, cattle and a few horses.

6. Other Animals recorded in the Catchment

There was one unconfirmed report of pig-rooting (Sus scrofa) in the forest on the true right of the Otira River below the mouth of the

Deception River. Pig-rooting was also reported in the head of

Thompson Creek (Poulter River) in the Waimakariri catchment.

Two stoats (Mustela erminea) were seen and a den found in the

roots of a totara at Kiwi Hut.

Two rats (Rattus rattus rattus) of the black colour phase were

caught in traps set in the forest at Kiwi Hut. Kernels of Miro berries

that had been opened by mice (Mus musculus) were found in the Townsend

Creek area. = 330 430 360 170 380 330 330 130 170 100 50 140 OPOSSUM 40 30 . 0 (J) 0 . {!) w 10 > ...c u ro RED DEER Q) 40 c ·~ (/) 1- 30 w _J ..:.J

> u HARE w 40 30 UJ 0:: LL 20 cf 10

0 r..--~~~~~~~~~~~~~-A-~ 1 . 3 4 5 6 7 7 8 g 10 11 1'2 Forest ssland VE GE TA TION ASSOCIA TION.S

KEY HIGH NIMAL USAGE

MEDIUM ANIMAL USAGE

L GE

Figure 6. Opossum, deer and hare interaction with the vegetation associations. ) 16.

ANIMAL - VEGETATION INTERACTION

Using the previous inference that a high pellet frequency is

synonomous with a high utilisation of the vegetation assocation by

an animal population, the following observations can be made.

(i) Deer - vegetation interaction

On the pellet evidence, the highest densities of deer occurred

in three forest associations - the bushline forest, short scrub

hardwood and the subalpine scrub, and two alpine grassland associations -

Chionochloa Eallens and Q• crassiuscula (Fig. 6B). Most of the pellets

found in the grassland were either in the vicinity of subalpine scrub

or in the tussock tongues that extended into the forest from the

head basins.

(ii) Opossum - vegetation interaction

The kamahi and rata-totara forest associations have distinctively

high pellet frequencies (Figure 6A). The bushline forest and alpine

scrub had low opossum pellet frequencies

A highly significant correlation was found between the frequency

of opossum pellets in plots and the percent browse damage on the

canopy trees (r = 0.813, P .01) (Fig. 7). Pellet density therefore

gives a good measure of the relative amount of damage caused by

opossums to the forest canopy.

Browse intensity appeared to depend largely on forest composition.

A great deal of browsing occurred in successional forest and in the mid

and lower altitude rata-kamahi-totara forest of the upper Taramakau and

Deception catchments. Much less browsing occurred in the beech

forests of the Otehake catchment. Low values were found for the

upper Otira catchment despite the abundance of preferred species.

The population may have been limited by trapping and control operations.

In rata-totara, kamahi, and successional forests the frequency of

browsed trees was greater than 30%, and they included most of the 17. dominant species (totara, rata, kamahi, wineberry, fuschia, and lace- bark). In beech forest the browsed trees were generally not canopy dominants, and the frequency of browsed trees was low.

(iii) Hare - vegetation interaction

Hares showed a distinct preference for the Chionochloa australis,

C. pallens and Celmisia species type of association, and the short vegetation in the vicinity of rock falls and screes (Figure 6c).

Hare pellets were very infrequent in the alpine scrub and in the

C. crassiuscula associations.

(iv) Animal - animal interaction

Taking into account the occurrence of pellets of the 3 species, deer, opossum and hares together, opossum prefer kamahi and rata totara compared with deer and hares; deer use bushline forest, short scrub hardwood and subalpine scrub more than opossums, and use

Chionochloa crassiuscula more intensively than hares. Strong overlap, with fairly high pellet densities, presumably indicates strong species interaction between deer and hares in the Podocaruus nivalis and

Chionochloa Ealle~~ associations (Fig. 6B and 6c).

CONDITION OF RED DEER

Challies (1969) found the deer in the Taramakau catchment to be

of equal (females) or larger (males) size than the population that

have been sampled in the Arawata or Okuru valleys. Both sexes were

smaller than animals in the Hokitika catchment. Thus, the Taramakau

population appears to be in fairly good condition, assuming that size

is indicative of the relative availability of food.

The kidney fat index data indicated that the Taramakau sample

was in poorer condition than a sample taken from the Hokitika catchment,

an area that has been under heavy control pressure.

Challies suggests that the Taramakau population is at such a

Figure Figure

51 51

.::----

5152 5152

8. 8.

53 53

Deer Deer

53545455 53545455

.,,..._ .,,..._

kills/man-day kills/man-day

5555 5555

555757585859 555757585859

for for

the the

NO NO

......

_ _

5960 5960 Taramakau Taramakau

6061 6061

6162 6162

block block

6263 6263

hunting hunting

~~~~~~~~~~~~~~~~ ~~~~~~~~~~~~~~~~

53 53

+ +

64656566 64656566

programme programme

676768 676768

...... 6869 6869

1·2 1·2

O· O·

1·0 1·0

50515152525353 50515152525353

Figure Figure

-'4!-WAGE -'4!-WAGE

9. 9.

Chamois Chamois

Nus-

545555 545555

kills/man-day kills/man-day

565757 565757

5858 5858

for for

5959 5959

NO NO

I I

the the

I I

OPERA OPERA

I I

608J 608J

I I

Taramakau Taramakau

I I

61616262 61616262

TIO TIO

I I

~ ~

WORK~~~~~~~~~ WORK~~~~~~~~~

block block

6363 6363

64646565666667676B 64646565666667676B

hunting hunting

progr~mme progr~mme

+ + + + 18. density that it has exerted heavy pressure on preferred food resources, and it is now utilising non-preferred food to a considerable degree.

PREVIOUS CONTROL PROGRAMMES

Tpe success of the control programme on the deer and chamois population can only be assessed accurately from 1955 to 1968, when the method of payment has remained the same.

Results from the control programme over the 13 years show a reasonably stable return for effort in kills per man-day per season

(Figures 8 and 9, Appendix 1). There is some indication of a lower yield over the past 4 years. This suggests that over the past 10 years the population is being maintained at a relatively constant density.

In the Otira area which is relatively intensively settled and is hunted by local residents, control operations appear to be keeping the population density low. This area, which contains short scrub hardwood and other favoured associations is probably capable of sustaining a

high deer density, although the pellet densities indicate a low

density population. The low density may be accredited to control

operations and to a lesser extent, to hunting for sport.

In the Otehake area the numbers of deer have probably been limited

by a unfavourable habitat, along with continued control operations. In

the lower reaches of the valley where access to shooters is good, deer

were confined to the forest, while in the upper reaches where access is

more limited the deer tend to utilise the grassland more than the

forest.

The control programme seems to have had little effect in lowering

the density of the deer population in the forest.and grassland of the

True Right, and the forest of the Deception and Top Taramakau areas.

In view of the past 4 years control operations (Appendices 1 and 2)

when comparing commercial helicopter and private hunting operations it 19. is obvious that helicopter recovery operations are the most effective form of control. Permit shooting directed from the National Parks

Board, and private hunting contribute very little to the overall deer control programme. From the results of this survey a reassessment of the control programme is needed.

DISCUSSION AND RECOMMENDATION FOR FUTURE CONTROL

1. Deer

Future control operations should be directed towards the areas where critical damage or modification to vegetation has occurred:-

(i) the short scrub hardwood associations.

(ii) kamahi forests.

(iii) Chionochloa australis/~. Rallens associations 0f the

Top Taramakau.

The Otira area has a good regeneration of plant cover and low

numbers of deer. The future control programme for this area could be

left more under the guidance of the Arthurs Pass National Park in

issuing shooting permits for this area.

In the Deception area an intensified control programme should be

initiated, aimed particularly at the forest areas where very large

deer pellet densities were recorded.

The vegetation cover of the Otehake river area is reasonably

stable and the average pellet density for the area is a little higher

than the Otira. A small increase in hunting intensity could be

justified in the upper reaches of the Otehake valley.

Hunting in the Taramakau catchment as a means of control is limited

in a number of ways. Evidence from the survey data has shown that

ground shooting has tended to move the deer into the upper altitude

forest and subalpine scrub belt, where shooting is more difficult.

This is distinctly evident when pellet data on relative use of forest 20.

and grassland in.the survey area are compared with data collected from

the Harper-Avoca catchment (Batcheler and Logan, 1963) and with the

vegetation assessment of the Takitimu Motintains (Evans, 1969).

The kills in the past have remained at about 2 deer per man-day.

Therefore for any effective increase in yield the number of man-days

per season would have to be increased.

Damage to vegetation in the catchment tends to be localised and

therefore control needs to be orientated towards these areas to obtain

the greatest return for effort expended. The recommendation for future

control operations is therefore to employ two techniques; helicopters,

and poisoning with 1080 on broadleaf.

The suggested programme is to completely close the area to hunting

until 1st February 1970 to allow the deer to move out. The grassland

from the forest making them more accessible to helicopter recovery

teams. Any shooting on the river flats and in the low altitude forest

should be allowed as this will tend to move the deer toward the grass-

land.

An outline of the 1080 poisoning programme is given in the summary

report of the survey (Wardle and Hayward, et. al., 1969).

2. Chamois

The highest chamois pellet density in the forest was in the

Otehake, while in the grassland high densities were recorded along

the Aickens and Barron ranges. The chamois population appears to be

at a manageable density. Chamois could probably be controlled more

economically if an incentive scheme was initiated for venison recovery ""- ·- -

teams to hunt chamois as well as red deer.

3. Opossums

Areas where high opossum pellet densities occurred and where

future control operations should be directed are the kamahi and rata-

totara forests of the True Right of the Taramakau between Hut and 21.

Alexander creeks, the Bald range , the head of Pfeifer creek, and the

Deception river. In these areas opossums are beginning to cause serious defoliation of the kamahi and rata-totara forest canopies.

James (1969) recorded that about 30% or more of the palatable trees are being moderately or severely browsed. This correlated with a 30% pellet frequency found in the sample plots. Therefore, opossum control needs more emphasis than it has had in the past. Three proposals for future control are:-

(i) exclusion of private fur collectors from the critical areas

where no interference in the control campaign should occur.

(ii) an intensive winter 1080 poisoning programme.

(iii) a population sample should be taken before and after the

poisoning to indicate the possible success of the poisoning pro

gramme using some condition index (see Bamford, 1968). The

exclsuion of private operators will enable a direct comparison

to be made between poisoning and the subsequent condition of the

opossum population.

The poisoning programme should be implemented as soon as possible,

and combined with current opossum population poisoning research under

taken at this station.

4. Hares

The highest density of hares occurred in the grassland of the Top

Taramakau area. This area, along with the Kelly range and Goat Hill)

is described by Evans (1969) as being in poor condition. 22.

REFERENCES

Bamford, J.M. 1968: Assessment of opossum poison campaigns, Westland,

1968. N.Z. For. Serv. UnEub. Report. November.

Batcheler, C.L. 1961: Report on the browsing and grazing mammals of

the Wairau catchment. N.Z. For. Serv. Permanent Report. No. 200.

1969: Animals, in Report on Vegetation, animals and

management (recommendations) for the Taramakau catchment.

N.Z. For. Serv. Prat. For. Branch Report. No. 69.

Caughley, G. 1963: Dispersal rates of several ungulates introduced

into New Zealand. Nature 200 (4903): 280-281.

Challies, C.N. 1969: Deer population in the Taramakau catchment.

N.Z. For. Serv. Unpub. ReJ25?..£!.

Christie, A.H.C. 1963: The ecology of the chamois (Rupicapra rupicapra)

in an alpine basin in southern Nelson. M.Sc. Thesisi Vic. Univ.

Wgtn., N.Z.

Donne, T.E. 1924: The Game Animals of New Zealand. John Murray,

London. 322 PP•

Evans, G.R. 1969 (a): The alpine grasslands of the Takitimu mountains.

N.Z. For. Serv. Prat. For. Brarch Report No. 54.

1969 (b): The alpine grasslands of the Taramakau catchment.

N.Z. For. Serv. Prat. For. Branch Report No. 61.

James, I. 1969: Opossum browse. N.Z. For. Serv. Unpub. Report.

Lambert, R.E. and Batcheler, C.L. 1969: Uses and limitations of pellet

counts in routine management and control of deer populations.

N.Z. For. Serv. Prat. For. Branch Reper~ No. 65.

Logan, P.C. and Harris, L.H. 1967: Introduction and establishment of

red deer in New Zealand. N.Z. For. Serv. Inf. Ser. No. 55.

and Mangos, B.G. 1953: Report on the Hurunui valley survey.

Unpubl. report DeEt· Int. Affairs. N.Z. Forest Service: Preliminary report and plan for control of noxious

animals in the Hokitika river catchment. Unpubl. report N.Z. For.

Serv.

Riney, T. 1955: Evaluating the condition of free-ranging red deer

(Cervus elaphus) with special reference to New Zealand.

N.Z. J. Sci. Tech. 36B: 429-463.

Thomson, G.M. 1922: The naturalisation of animals and plants in

New Zealand.

Cant. Univ. Press. 607 pp.

Wardle, J.A. and Hayward, J.D. 1969: The forests and scrublands of

the Taramakau.

N.Z. For. Serv. Prat. For. Branch Report No. 59. APPENDIX 1 - Summer Season Animal Kills. Government O;eerations Taramakau Block

Season Deer Deer/ Chamois Chamois/ Payment Man-day Man-da;z Method

1950/51 325 2.56 51 0.16 Wage-bonus 1951/52 789 2.51 60 0.19 II II 1952/53 738 2.01 56 0.15 II II 1953/54 660 1.77 26 0.07 II II 1954/55 991 2.84 38 0.15 II II 1955/56 1278 4.15 139 o.45 Piecework 1956/57 1635 5.21 191 0.61 11 1957/58 527 1.85 77 0.27 II 1958/59 810 2.62 97 0.31 II No shooting operations 1959 to 1961 1961/62 1203 2.12 691 1.22 " 1962/63 * 889 1.34 340 0.51 II 1963/64 1248 3.32 174· o.46 " 1964/65 65·1 2.09 57 0.18 II 1965/66 323 1. 90 157 0.92 II 1966/67 466 1.57 53 0.18 It 1967/68 548 1.99 188 o.68 11 1969 ** 147 2.45

* Taipo and Taramakau blocks combined

**FE-.:RES autopsy shooting during survey. APPENDIX 2 - Animals killed, National Parks Board permits and Commercial Venison Recovery - Taramakau Catchment

Year No. of permits Locality No. deer No. Chamois a) Weekend Permits, National Parks Board

1968 54 Oth~a-Deception 16 8 19691 25 II II 13 7 b) Six monthly Permits, National Parks Board 1966 21 Otira-Deception 41 13 1966 18 II II 52 8 1967 19 II II 47 11 19673 5 Taramakau River2 274 27 19684 2 II II 1968 10 Otira-Deception 55 7 c) Commercial Hunting (private) 1966/67 Taramakau River 200 1967/68 II II 250 1968/69 " II 40 (Season up to March 1969) d) Commercial Hunting (helicopter) June/July 1967 Alexander Range 302 II 11 II True Left Taramakau 118 Sept. 1967 Alexander l~ange 76 Nov. 1968 !I II 147 Dec. 1968 II II 183 Jan. 1969 Taramakau Upper 25 Feb. 1969 It II 76 March/April 1969 True Left Taramakau 92 II II II True Right II 13

1 1969 returns up until March, 1969. 2 Most of the deer in this tally were shot for carcass value, see Table 2 (c).

3 Some 1968 returns are still outstanding. 4 Most of the 1967/68 and some of the 1968/69 animals are recorded in the National Parks Board six monthly permit for 1967. APPENDIX 3 - Description of Dominant Specie~ in the Vegetation Associations

Association Number 1. Red Beech 2. Kamahi 3. Rata - Totara 4. Mountain - Silver beech 5. Bushline forest 6. Short scrub hardwood 7. Subalpine scrub - this association included plots described by the forest survey (above) and those grassland plots that included Dracophyllum species and Phormium colensoi associated with alpine grasses

8. ~odoca~rus nivalis dominant with Chionochloa pallens, Q· flavescens and Celmisia species.

9. Chionochloa J2~11ens dominant with Celmisia species and Poa species.

10. C. austra1i~ dominant with C. pallens and Celmisia species. 11. C. crassiuscu1a with _Q. pallens, Paa and Celmisia species. 12. Rock and Scree - mostly rock falls or scree field with short vegetation at the edges or scattered throughout the screes.