untitled

country. Table 20.2 lists the sequence of official justifications for government-funded

control of deer from 1920 onwards (Caughley 1983).

Whereas the stated justification for the control operations changed with time, those

changes had virtually no effect upon the management action. There were certainly

changes in control techniques but, with the exception of the change of 1967, these

were evolutionary adjustments in the management action itself. They were not

driven by changes in policy. The means themselves were the end.

Up until 1980 the reasons given for the control operations were that deer and other

species caused erosion of the higher slopes and silting of lower rivers (Table 20.2).

However, in 1978, new meteorological, hydrological, geomorphological, and

stratigraphic research showed that deer, chamois (Rupicapra rupicapra), and tahr

(Hemitragus jemlahicus) had little or no effect on the rate at which river beds silted

up or on the frequency and size of floods. Despite these data, deer control continued

after 1980 for no verifiable reason. All that changed were the stated objectives, which

were variously for “aesthetics,” for “proper land use,” to “ensure the continuing health

of the forest,” to “protect intrinsic natural values,” and to “maintain the distinctive

New Zealand character of our landscapes.” These are not open to scientific testing.

Many similar examples could be cited from other countries. Control operations

must have clear objectives framed in terms of damage mitigation. Their success must

be measured by how closely those objectives are met, not by the number of animals

killed. The operations must be costed carefully to ensure that their benefit exceeds

their cost. And their success or failure must be capable of independent verification.

Table 1.1 gives a matrix of possible objectives and actions. It can be filled in to ensure

that the management action is appropriate to the chosen objective.

There are three circumstances in which control may be an inappropriate manage-

ment action: (i) where the cost exceeds the benefit; (ii) where the “pest” is not in

fact the cause of the perceived problem; and (iii) where the control has an unac-

ceptable effect upon non-target species. These are best investigated experimentally

before a control program is instituted. We give two examples.

Cats were introduced to the subantarctic Marion Island in 1947 to deal with house

mice marooned by shipwrecks. They increased rapidly to 3000 by 1977 and fed mostly

on ground-nesting petrels. The breeding success of the petrels, particularly the great-

winged petrel (Pterodroma macroptera), seemed to be declining and cats were sus-

pected to be the cause. The neighboring island of Prince Edward was conveniently

free of cats and became the experimental control. The objective of reducing cats was

to increase the breeding success of the petrels. Hence the success must be defined

in terms of the birds’ breeding success, not in terms of reduced numbers of cats. An

WILDLIFE CONTROL 357

Years Official objectives of deer control

1920 –29 Increase the size of antlers

1930 –31 Reduce competition with sheep

1932– 66 Prevent accelerated erosion generally

1967– 80 Prevent accelerated erosion in the heads of rivers that may flood cities

1981–92 No verifiable reason offered

From Annual Reports of Department of Internal Affairs and New Zealand Forest Service.

Table 20.2Published
official justification for
government control
operations against deer
in New Zealand.

20.5 Determining whether control is appropriate