Other formats

    Adobe Portable Document Format file (facsimile images)   TEI XML file   ePub eBook file  


    mail icontwitter iconBlogspot iconrss icon

Proceedings of the First Symposium on Marsupials in New Zealand

Damage By Possums Trichosurus Vulpecula To Farm Crops And Pasture

page 197

Damage By Possums Trichosurus Vulpecula To Farm Crops And Pasture

* Present address: Wildlife Service, Department of Internal Affairs, Private Bag, Wellington.


This paper reviews previous reports of possum damage to crops and pasture, methods of measuring the damage, and the cost-effectiveness of control operations.

Various reports of possum damage indicate that localised serious damage can occur to crops (e.g. turnips, chou moellier, lucerne) and to pasture (clovers and grasses), particularly near forest-scrub margins.

Methods previously used to assess damage (questionnaire surveys and analysis of possum diets) are incomplete and possibly biased. The questionnaire surveys used in the past have not attempted to estimate losses, but only the type and extent of the damage. If incomplete consumption or compensatory growth occurs, loss of farm production cannot be directly extrapolated from analysis of possum diets. Assessment of economic losses must be based on measurements of yield.

Current methods used to control possums (e.g. phosphorus and 1080 poisons) are adequate to reduce possum populations locally. However, achievement of satisfactory control is limited by a lack of co-operation and co-ordination as well as shortage of finance. Research cannot be expected to solve these problems, but there is a need for research into cost/benefits and methods of improving the efficiency of control.


If control of the common brushtail possum Trichosurus vulpecula on farm land is to be economic, the cost of control should not exceed the value of the damage (reduction in yield) caused by possums. Thus, assessment of the economic impact of possums is a pre-requisite for economic control. This was recognised by the Technical Advisory Committee (Animal Pests), which stated in 1969 that "while there is no evidence that opossums compete seriously with livestock for pasture on a wide scale, locally opossum grazing can appear serious enough for farmers to persuade pest destruction boards to carry out intensive control. The effects of opossums on pasture should therefore be investigated before such intensive control can be justified" (Anon 1969). This is even more crucial today because of the ever-increasing costs of control operations.

As a preliminary to any cost/benefit analysis of possum control on agricultural land, it is necessary to devise methods of measuring the damage (ultimately in terms of the market value of the units lost) and the cost-effectiveness of control. In this presentation we give:

page 198
1.A brief review of previous reports of possum damage to farm crops and pasture;
2.An outline of the methods and problems of measuring the damage; and,
3.An outline of the problems of assessing the cost-effectiveness of control.

Previous Reports of Damage

Early reports on the feeding of possums were based on observations of possums in indigenous forest. They concluded that "opossums are not grass eating animals" (Mason 1958)*, and that "the opossum is not and will never become an important grassland pest" (Howard 1963). The diet of possums in pastoral districts was first analysed by Gilmore (1965a, 1965b, 1967). He found that "where root crops were grown as winter feed for stock opossums travelled long distances to feed on them" (Gilmore 1965b), and that "clover and grasses were eaten in quantity and by large numbers of opossums throughout the year" (Gilmore 1967). He speculated that "opossums may seriously reduce the carrying capacity of the land" (Gilmore 1965a). Shortly after that, Quinn (1968) reported an instance where the carrying capacity of one farm had been raised by about 1 ewe/hectare following an intensive poison campaign against the possums, but he also stated that this case "was possibly exceptional". About the same time, surveys (consisting of questionnaires supplemented by interviews and inspections) in several agricultural districts (Anon 1966, 1967a, 1967b, 1968) all indicated that possum damage to crops and pasture was of minor importance. This is illustrated by two quotations: "crops close to forest fringes were damaged to some extent but, taking cropping as a whole, the damage was not significant" (Anon 1967b) and "opossums do eat limited amounts of pasture … but … there is no evidence that they have affected the carrying capacity of the land generally" (Anon 1967b). More recently, Harvie (1973) published the results of stomach analyses of possums collected in 1968 and 1969 from four farms including the one reported on by Quinn (1968). She estimated that pasture formed about 30 percent of the diet of possums sampled while feeding in paddocks, and, using a figure of 43 possums/ha, calculated that the stocking rate could possibly have been increased by 2.5 sheep/ha. (Recalculation of these data indicate an increased stocking rate of 1.3 sheep/ha: Fitzgerald 1977).

From our preliminary investigations, it is evident that serious localised page 199 damage can occur, particularly near forest or scrub margins. The damage to pasture and crops (such as turnips, chou moellier, lucerne) is most noticeable when paddocks are shut-up without stock (e.g. pasture shut up in autumn for winter feed).

* Thomson (1922) reported that "the opossum is not a grass-eating animal".

Measurement of Damage

The previous investigations have approached the problem of measuring possum damage in two ways: (i) by questionnaire surveys, and (ii) by analysis of stomach or faecal contents coupled with estimates of the density of possums. Both approaches are indirect and both contain errors. The results of questionnaires are possibly biassed from a variety of sources of error (Deming 1944, Oppenheim 1966). For example, farmers are not always aware of, or concerned about, possum damage on their properties, but this may not be apparent from their response to questionnaires. However, provided its limitations are recognised, a questionnaire survey can give a general appraisal of the situation which might serve as a guide to policy. Quantitative stomach analyses are also subject to several sources of error, including the difficulty of estimating the extent of root and tuber consumption (Vorhies and Taylor 1933), and faecal analyses require accurate determination of the differential break-down rates of plant materials (e.g., Dunnet et al. 1973, Fitzgerald 1976). However, the most serious limitations in extrapolating from analysis of diet to loss of production are: (1) the strong dependence on accurate estimates of the density of possums feeding on an area of farm land, (2) no estimate is made of any compensatory growth by plants in response to browsing or grazing, and (3) no estimate is made of the amount destroyed but not eaten. In such cases, reliable estimates of losses to the farmer must be based on measurements of yield (various references in Chiarappa 1971).

Methods of assessing reduction in yield caused by pests have many similarities to agronomic and horticultural yield assessments, such as described by Boswell (1974).

We adapted one of these methods to estimate losses to possums, as an initial step in a cost/benefit analysis, by placing wire-netting cages in unstocked paddocks to exclude possums from small replicated plots of about 3 square metres. Each protected plot was paired with an unprotected plot (separated by only 1 m to minimise site differences), and a paired comparison t-test (Sokal & Rohlf 1969) was used to compare yield. In only one trial out of six was there a significant reduction of yield definitely attributable to page 200 possums. This was a 26% reduction in a chou moellier/swede crop, but because of the small area involved (0.4 ha), the crop lost to possums in 3 months would have supported only 336 ewe equivalents for 1 day. It was felt that these results reflect the limitations of the trials rather than a lack of significant possum damage.

The method we used has the following disadvantages:

1.It can only be used over a small area at a time.
2.It is costly, time and labour-consuming.
3.The protected (enclosed) plots probably have a modified micro-climate which may influence yield (see Lynch 1966), although if this is predictable it may be possible to correct for it.

However, it is possibly the only method suitable for use on pasture.

For crops it would be desirable to develop an alternative method to give a measure of the loss after the damage has occurred. This could perhaps be achieved by scoring the frequency of damage, and determining the dry weight yield of damaged and undamaged plants at points along a transect. Having established a quantitative relationship between the amount of damage and the resulting loss of production, it should be possible to survey the amount of damage to estimate losses on a regional basis.


Control of possums, when declared a pest on rateable land, is the responsibility of local Pest Destruction Boards. These Boards are elected by local ratepayers, and they employ staff to carry out control operations. Finance for this work normally comes from rates from farmers (and a 1:1 government subsidy).

Methods used to control possums on farmland are many and varied (Anon 1966). Probably the most common methods are the use of phosphorus (jam) and 108Q (jam and carrots). 1080 is probably the most effective but is not favoured in certain areas because of the risk of poisoning domestic animals, particularly farm dogs, from secondary poisoning. Individual Board operators have usually evolved techniques applicable to their particular area. Wherever possible, they make use of local features, such as laying poison around pine trees in spring or feed crops (e.g. turnips) in autumn/winter. The need to maintain full-time employment of staff means that seasonal use is often made of page 201 several techniques and that staff are usually involved in rabbit Oryctolagus cuniculus control as well. To estimate the cost of this control, it would be necessary to separate the various costs in Pest Destruction Board records.

Existing techniques used by Pest Destruction Boards appear to be adequate to reduce local possum populations. Examples of successful control operations can be cited in many Board areas. However, satisfactory control is seldom achieved throughout a Board's area. There are two basic reasons for this:


Insufficient co-ordination of activities.

Breakdowns in communication have, particularly in the past, resulted in failure or only poor success of control, e.g. when Boards have not given farmers sufficient notice to move stock, or when farmers have refused access to Board operators for various reasons, including trapping for skins, and their inability to de-stock some areas requiring control. For successful control it is essential that good co-operation exists between Boards and ratepayers. The Agricultural Pests Destruction Council has recognised this and is actively endeavouring to bridge the gap (e.g. see A.P.D.C. Annual Report 1977). It is essential that the communication be two-way, and that farmers, for example, advise the Board of any cropping and pasture development programmes.


Lack of finance.

Pest Destruction Boards are dependent on rates from farmers. However, the Boards are farmer-controlled, and in some cases appear reluctant to increase rate demands on fellow farmers, particularly when areas without a possum problem are included within the Board's jurisdiction. Few Boards strike differential rates. Most Boards are also reluctant to lower the level of rabbit control, which would make more money available for control of possums, for fear that rabbit populations might explode in the absence of continual control. In many cases, possums are causing more damage than rabbits, but the rabbits are still being hunted whereas the possums are not. There would appear to be a need for some form of differential rating or payment-for-service-rendered scheme.

These organisational and financial difficulties will need to be accounted for in any assessment of the cost-effectiveness of possum control.

page 202

Concluding Remarks

Previous reports and our own experience all indicate a situation typified by serious damage of a localised rather than a generalised nature. There is a definite need for research to determine:

1.The economic importance of this damage (on a local and regional basis).
2.The short and long term cost/benefits of current control practices.
3.The areas of improvement to control technology, either by attempting to increase efficiency or by development of new methods.

Research cannot be expected to solve problems associated with the organisation and financing of pest control.


Agricultural Pests Destruction Council 1977. 9th Annual Report 1976–1977.

Anon. 1966. Departmental report on opossum damage in the East Coast Rabbit District 1965–1966 . Inter-Departmental Report (unpublished).

Anon. 1967a. Investigation of proposed opossum control plan (Tinui Rabbit District). Inter-Departmental Report (unpublished).

Anon. 1967b. Report on opossum control in New Zealand. Inter-Departmental Report (unpublished).

Anon. 1968. Report on King Country opossum survey. Inter-Departmental Report (unpublished).

Anon. 1969. Report and recommendations for additional research into control of animal pests. Technical Advisory Committee (Animal Pests) Report (unpublished).

Boswell, C.C. 1974. Pasture measurement. Ministry of Agriculture and Fisheries, Invermay Research Centre , Technical Report No. 2.

Chiarappa, L. (ed), 1971. Crop loss assessment methods. F.A.O. manual on the evaluation and prevention of losses by pests, disease and weeds.

Deming, W.E. 1944. On errors in surveys. American Sociological Review 9: 359–69.

Dunnet, G.M. , Harvie, A.E. & Smit, T.J. 1973. Estimating the proportions of various leaves in the diet of the opossum, Trichosurus vulpecula Kerr, by faecal analysis. Journal of Applied Ecology 10: 737–45.

Fitzgerald, A.E. 1976. Diet of the opossum Trichosurus vulpecula (Kerr) in the Orongorongo Valley, Wellington, New Zealand, in relation to food-plant availability. N.Z. Journal of Zoology 3: 399–419.

Fitzgerald, A.E. 1977. Letter to the editor. Proceedings of the N.Z. Ecological Society 24: 139.

page 203

Gilmore, D.P. 1965a. Opossums eat pasture. N.Z. Journal of Agriculture 110: 284–86.

Gilmore, D.P. 1965b. Food of the opossum Trichosurus vulpecula in pastoral areas of Banks Peninsula, Canterbury. Proceedings of the N.Z. Ecological Society 12: 10–13.

Gilmore, D.P. 1967. Foods of the Australian opossum (Trichosurus vulpecula Kerr) on Banks Peninsula, Canterbury, and a comparison with other selected areas. N.Z. Journal of Science 10: 235–79.

Harvie, A.E. 1973. Diet of the opossum (Trichosurus vulpecula Kerr) on farmaland north-east of Waverley, New Zealand. Proceedings of the N.Z. Ecological Society 20: 48–52.

Howard, W.E. 1963. Rabbit control - time for a change? N.Z. Farmer 83(27): 2–5.

Lynch, P.B. 1966. Conduct of field experiments. Department of Agriculture, Bulletin No. 399.

Mason, R. 1958. Foods of the Australian opossum (Trichosurus vulpecula, Kerr) in New Zealand indigenous forest in the Orongorongo Valley, Wellington. N.Z. Journal of Science 1: 590–613.

Oppenheim, A.N. 1966. Questionnaire design and attitude measurement. Heinemann, London.

Quinn, J. 1968. Opossums deplete Wanganui pastures. N.Z. Journal of Agriculture 117: 17–20.

Sokal, R.R. & Rohlf, F.J. 1969. Biometry. Freeman & Co. San Francisco.

Thomson, G.M. 1922. The naturalisation of animals and plants in New Zealand. Cambridge University Press.

Vorhies, C.T. & Taylor, W.P. 1933. The life histories and ecology of jack rabbits Lepus alleni and Lepus californicus sp. Arizona Agricultural Experimental Station, Technical Bulletin 49: 471–587.

General Discussion

B.D. BELL. You mentioned that control of possums on rateable pastoral land is the responsibility of the local Pest Destruction Board. Does the Board have powers to enforce possum control if it is against the wish of the farmer?

NELSON. When the possum is declared a pest of local importance then the local Board has legal right of access for control. However they generally liase with the landowner to ensure control operations phase in with his requirements and to ensure stock losses are avoided. In the interests of furthering good relations they are reluctant to push their enforcing powers.

page 204