The Adaptation and Increase in the Paradise Shelduck (Tadorna variegata) Within a Man-modified Environment

Transactions of the Royal Society of New Zealand : Zoology, Volume 6, Issue 12, 6 August 1965, Page 115

The Adaptation and Increase in the Paradise Shelduck (Tadorna variegata ) Within a Man-modified Environment

H. J. F. McAllum

By

[Received by the Editor, June 22, 1964.}

Abstract

The colonisation of the study area by the Paradise Shelduck is related to the reduction of forest, an increase in pastureland and in the number of stock ponds.

Critical factors for breeding, feeding and moulting sites are suggested and discussed in relation to the North Island distribution.

Introduction

Investigations were carried out from November 1960 to May 1962 to examine the habitat requirements of the Paradise Shelduck with a view to explaining the extension in range over the last 80 years from 39° S. to S. latitude.

The Paradise was unknown west of the Ruahine and Kaimanawa Ranges prior to 1908, but in that year a pair of birds brought from the South Island was liberated about three miles from National Park by Mr J. Cullen. Subsequent liberations occurred intermittently up to 1919. Eventually the birds became established in the Kaimanawa Range in the east of the study area. After about 1937 a large increase in the population became evident, the birds spreading from the Kaimanawas to the west to the developing pastoral land, denuded of forest.

Environment

Study Area. Because of the increase in the Paradise population at the southern end of the central upland plateau it was selected as the study area (Fig. 1). The approximate boundaries were, in the west, the Wanganui River; in the south, a line running from Wanganui through Hunterville to the Ruahine Range; in the east, the Kaweka and Ruahine Ranges; in the north, a line running east-west through Mt Ruapehu.

A large proportion of the time was spent in the area between Turakina Valley and the Kaweka Range.

Climate. The area excluding the Ranges rises to approximately 2,000 feet above sea level, and the prevailing westerly winds largely influence the cool superhumid climate. A large diurnal variation occurs, particularly in the more elevated regions, and the area is well watered with the maximum precipitation occurring in winter (Gamier 1958, p. 128). The country near Mt Ruapehu has a rainfall of 50-60 inches and the other areas 35-50 inches.

The climatic data for Karioi (2,125 ft) are representative of the area. The degree of variation occurring in other places is mainly associated with altitude. The data for Wanganui, a coastal city, indicate the range of variation within the study area.

(By courtesy of the N.Z. Meteorological Service.)

Soils and Vegetation., The soil is derived mainly from the Tongariro ash beds and originally supported a podocarp-broadleaf forest, partly destroyed by the Taupo eruption in recent geological times. Scrub and tussock have regrown on these areas and rain forest still exists on mountain regions, but a large area of tussock land has been devoid of forest for many centuries. The distribution of forest for 1880 is shown in Fig. 11.

The milling of the forest began on the completion of the main trunk railway in 1908, mainly in the vicinity of Raetihi and Ohakune, and the cleared areas were developed for farming. Land fertility increased, particularly with the aid of aerial topdressing, with a consequent increase in the number of sheep ana cattle (Farming in New Zealand, 1950, p. 140).

The pasturelands can be divided generally into first, second and third grades (modified from the Agricultural Department’s Pasture map, 1939).

First grade—ryegrass and white clover dominants, cocksfoot and paspalum.

Second grade—browntop dominant, sweet vernal, fog, suckling clover, lotus, dogstail and some danthonia.

Third grade—danthonia dominant, some hair grass, ratstail, rice grass and suckling clover.

Topdressing encourages the subterranean clover, ryegrass, cocksfoot, and white clover, particularly where oversowing is carried out with it. The study area is generally of this improved second grade pasture.

Rivers. The main rivers in the study area draining the central upland plateau are the Wanganui, Mangawhero, Whangaehu, Turakina and the Rangitikei. Grassy banks and intermittent flood plains are found interspersed with scrub and forest remnants along all the rivers, but the larger and swifter Wanganui and Rangitikei rivers have extensive gorges as well.

Natural and Artificial Ponds. The mudstone in the Turakina Valley and Taihape areas is prone to slumping, causing hummocks amongst which small lakes have formed. Other natural ponds form in depressions and in low-lying country from streams cut off by slips.

The increased use of the bulldozer after the 1939-45 war has contributed greatly to the construction of stock ponds. The porous soil of the Waimarino area, however, is not suitable for dam construction, whereas the stock ponds of the Hunterville-Taihape area are an important source of water (A. A. Duncan, pers. com.). The growth in the number of the ponds for the two areas is indicated in Table 11.

Methods

Distribution was determined by sight records and by examination of published records; changes in population from published records, conversation with old identities in the districts, and records of liberations. These were correlated with the decrease in forest area and land development.

Due to the nature of the country in the study area it was not possible to visit every part. Some of the steeper and more inaccessible country was viewed from the air and vegetation surrounding the ponds was recorded. When possible these were later visited on foot and measured. A total of 400 ponds was visited. The size of ponds carrying broods, distance from nearest tree, type of vegetation in and around the pond, age and number in brood were recorded.

Moulting, post-moulting and non-breeding flocks were recorded for numbers, sex ratios, type of pasture on which they were feeding, the size of the nearest pond and the distance to it.

Feeding

The goose-like trait of grazing exhibited by Paradise is evident when large flocks congregate after moulting. They are reputed to damage pasture by pulling grass, particularly in new pastures or crops and by fouling it. This was recorded as early as 1871 by Travers (p. 207).

The density of birds per acre varies with the time of the year. Table 111 shows the order of density per acre for the larger autumn post-moulting flocks and the smaller spring non-breeding flocks. The areas over which the birds were feeding at a particular time were paced out, converted and corrected to the nearest half-acre. Single birds apart from the main feeding flocks were discounted, but the flocks do not remain static, moving in several directions at once. If disturbed the birds may fly to other parts of the paddock or to another paddock, but if undisturbed the area grazed seems to be not much greater than the area they were first observed feeding on.

The spring non-breeding flocks are often found on or near the areas where the autumn post-moulting flocks feed, as well as elsewhere. The density is not as great, the smaller flocks feeding over a relatively greater area than post-moult-ing flocks. As Table IV shows, the post-moulting flocks are very close to water, generally a stock pond. Feeding takes place not only on the best paddocks in the vicinity, but on the greener parts, which are usually near the pond or its outflow.

At times of drought the density of birds becomes much higher around the stock ponds with water.

It is suggested that the requirements of these flocks are food, preferably young green herbaceous growth, water and lack of disturbance.

Breeding

Broods first appeared on stock ponds in September. The size of the ponds with broods varied from 108 sq yds to 3% acres. The average size for ponds with and without broods was not significantly different. The areas of the ponds were calculated from the appropriate formulae and measurements were paced out and converted to yards.

The distance of ponds from the nearest stand of trees was measured by pacing and Table V shows that the majority of these were within 100 yards. Ponds not supporting broods show a tendency to be further from stands of trees than those with broods.

Comparison of ponds was made for overhead cover, invertebrate fauna and vegetation in and around the ponds but no significant difference was found between those supporting and those not supporting broods. The birds do not appear to require overhead cover around the edge as 11 of the 51 ponds with broods had clean pasture edges.

Ponds completely surrounded by bush to the edges did not support a brood.

In the first few weeks the ducklings when disturbed swim to the middle of the pond and dive. The young when fully feathered (except for the primary feathers) prefer, when disturbed on the smaller ponds, to hide in cover at the pond’s edge but if cover is lacking, they depend on diving. Hiding occurred when the parents gave the first danger call and did not depend on the distance of the source of disturbance from the pond. When the young could see the source of disturbance they did not attempt to hide. Once in hiding the young birds remained motionless and could be picked up by hand.

The adult birds are known for their aggressive behaviour when defending a territory or a brood. (Delacour 1954, p. 246; Wilkinson 1931, p. 88; Henry 1907, p. 171.) In addition to this the ponds probably supply a protected brooding area and brood survival is high. This is suggested by the brood survival for 22 broods at the flying stage, the average being 6 and the range from 3 to 10 per brood. Only 1 brood was seen on a river although pairs were found along the banks at intervals.

Within the study area the six nests found were all associated with trees. The habitats of the nests were;

(i) On the ground within a tall hollow tree stump at the edge of a stand of native bush and 196 yards from water.

(ii) Within a split of a decaying log in an open paddock and 450 ■ yards from water;

(iii) In an epiphytic Astelia sp. 15 feet above ground in a Matipo tree [Myrsine sp.) and 40 yards from nearest pond.

(iv) In an open hollowed out cabbage tree (Cordyline australis) bole 97 yards from water.

(v) In an Astelia in a beech tree (Nothofagus sp.) about 40 feet above ground at the edge of a large stand of native bush and about \ mile from water.

(vi) In an Astelia in a Maire tree (Mida salicifolia ) about 20 feet above ground and 90 yards from water.

Four of the six nests were within 200 yards of water and all within 450 yards.

Two other nests are on record for the North Island, one at Lake Waahi, Huntly, in tussocks {Carex sp.) 200 yards from water and the other at Masterton, in a hole in a bank about 7 feet directly above water.

Moulting

During late December, January and early February, the Paradise moult their flight feathers. An individual bird is flightless for about two weeks.

Ten moulting sites were known in the study area. The area of water varied from about 0.5 acres to 20 acres and supported from 50 to 900 birds or more at any one time.

Of the 10 sites, nine have pasture around all or part of their edges, three have raupo (Typha meulleri ) along the major length of their edge, another three have raupo present, five have trees close to their edge, all have a slope or hillside adjacent and are greater in depth than required by a bird swimming under water, a minimum of 18 inches to 2 feet. The hillside allows the birds to maintain a careful lookout while feeding or resting for any sign of disturbance and it also aids in their return to the water. The birds form raffs on the water and the continual movement allows for estimation only of numbers. In January 1962, 800-900 birds were in a raft on acres of water and 40-50 on acres. Other sites may occur in the study area where only a few birds moult on each.

Summary of Critical Factors for Feeding, Breeding and Moulting

It may he suggested tentatively that the following conditions are critical: (i) Young herbaceous growth in close proximity to water for preferred feeding areas.

(ii) Preferred nest sites near suitable water for brooding. Cover is not essential for brooding area.

(iii) For moulting, large areas of water with sufficient depth for diving, access to the edge for resting and feeding, and to a hillside.

Discussion

It can be seen from Figs. 1 and 2 that a great reduction in forest has occurred, it being replaced by pastureland. The environment of the study area at the time of the first liberations was beginning to be changed. The liberated birds having no tradition for the liberation sites, left and found a preferred habitat in the Kaimanawas, particularly it seems on the river flats of the Rangitikei River. Early reports of birds seen near Mt Ruapehu (W. S. Strachan, T. Shout, pers. com.) suggest they travelled about in the area. However, Mr R. Fraser (pers. com.) cannot recall birds in the upper reaches of the Ngaruroro River or the Ngamatea swamp in 1919 or 1937. They were, however, present there in 1938-39 (N. Elder, B. A. Vercoe, pers. com.) —i.e., at the same time as the noticeable population increase. There is no evidence to support the theory of an influx of birds from the east coast into the upper Ngaruroro River, Ngamatea swamp or Rangitikei River, prior to invading the study area.

The increased number of stock ponds associated with the improvement of pasture and increased stock density have provided the Paradise with suitable brooding and moulting sites.

The nest sites in the study area were all associated with remnant stands of native forest. In many cases grazing animals wandering through the stands have suppressed regeneration and consequently these will, in the course of time, die, decay and disappear. The optimum habitat may be transitory, for disappearance or reduction in the forest remnants could become a limiting factor.

Distribution

It is evident from a comparison of Figs. 2 and 3 that the extension of range has been from the south, northwards.

Buller (1888, p. 265) described the distribution as “abundant in the Wairarapa and the Ruataniwha plains, but they do not extend above 39° S.” He also recorded them as scarce on the west coast of the North Island.

Kirk (1895, p. 11) wrote, “About 15 years ago the Paradise duck was very common on the East Coast of the Wellington District between Gape Palliser and Gastlepoint, but at the present time the traveller may ride the entire distance without seeing a specimen The present position in the Wairarapa seems to be only a little better than that described by Kirk.

The northerly trend is also recorded by Guthrie-Smith in his book Tutira (p. 318) where, prior to 1917 the Paradise was a rare visitor, but in that year they bred and reared broods. He suggested the reason for this in “The three species that have been attracted to the station by the changed conditions are the Banded Dottrel ( Charadrius bicinctus), the Pied Stilt ( Himantopus leucocephalus) and the Paradise Duck ( Casarca variegata)”.

Agriculture has changed the Gisborne-East Cape area from one of predominantly forest to one of predominantly pastureland. In the spring of 1943 a large number of Paradise invaded the Gisborne district and then spread northwards where they are now well established. (A. Blackburn, G. Hair, pers. com.)

Within the Rotorua district small numbers of birds have been recorded for many years.

Nesting is known to have occurred in recent years (1960-61) at Lakes Waahi and Whangape (R. Adams, pers. com.) both of which are close to the 37.5° S. latitude levels. Above 37.5° S. the sightings are probably only of visitors, but it can be expected for the birds to breed above this and build a tradition for the areas as the conditions stated previously come about with land development.

The sight records plotted in Fig. 11l have been recorded in the N.Z. Ornithological Society’s journal “ Notornis ”, by Wildlife Branch members and the writer.

Establishment of the Paradise within the study area may provide a source of supply for the gradual establishment in the Taranaki district aided by liberations there. It may also act as a source for the invasion of the birds into the developing country north-west of Mt Ruapehu.

To a large extent the present distribution conforms with second and third grade pastureland. Absence from first grade pastureland may be due to higher density in human population and therefore disturbance and a lack of suitable nesting habitat.

Acknowledgments

I wish to thank Dr K. E. Westerskov, Messrs T. A. Caithness and P. Morrison for helpful discussions and suggestions; Messrs K. H. Miers, I, A, E. Atkinson and D. Stenhouse for critically reading the manuscript; the Director, Department of Agriculture, for allowing me to use the Pasture (1939) map; the Editor of “Notornis” for allowing me to use the published sight records; Mr L. McMillan, of the Wellington Acclimatisation Society, for his help in the field; Mr A. A. Duncan for the estimates of stock ponds; the many farmers who have given generous aid, the field staff of the Wildlife Branch for their interest and valuable help, and particularly Mrs H. Hall.

References

Buller, W. L., 1888. A History of the Birds of New Zealand. 2nd Edition, London, pp. 264-267.

Delacour, Jean, 1956. The Waterfowl of the World. Vol. 1, Country Life Ltd., London, p. 246.

Dept, of Agriculture, 1950. Farming in New Zealand, pp. 209-222.

Garnier, 8., 1958. The Climate of New Zealand. Arnold, pp. 126-131.

Guthrie-Smith, H., 1921. Tutira. Blackwood. London, p. 318.

Henry, R., 1907. Emu. Vol. 6, p. 171.

Kirk, T., 1895. Displacement of Species in New Zealand. Trans. N.Z. Inst. 28: pp. 1-27. Oliver, W. R. 8., 1936. The Paradise Duck. Emu, Vol. 36, p. 69.

Soil Bureau, 1954. General Survey of the Soils of North Island, New Zealand. Bulletin No. 5.

Travers, W. T. L., 1871. Notes of the Habits of Some of the Birds of New Zealand. . Trans. N.Z. Inst. 4, pp. 206-213.

Wilkinson, A. A., and Amy, 1952. Kapiti, Bird Sanctuary. Masterton Printing Company, pp. 87-91.

H. J. F. McAllum, Veterinary School, Massey University of the Manawatu, Palmerston North.

Karioi Wanganui (1927-1960) (1937-1960) Average daily maximum daily maximum59° F. 59° F.63° F. 63° F. Average monthly maximum 80° F. 83° F. Extreme maximum maximum86° F. 86° F.88° F. 88° F. Average daily minimum daily minimum39° F. 39° F.49° F. 49° F. Average monthly minimum 19° F. 30° F. Extreme minimum minimum14° F. 14° F.29° F. 29° F. Average annual rainfall annual rainfall48 inches 48 inches34 inches 34 inches

Table I.—CLIMATIC DATA

Natural Artificial Year Ponds Ponds Waimarino County 1908 50 — 1961 50 100 Rangitikei County 1908 350 — 1961 350 2,000 Estimated totals for 1961 400 2,100 (These -estimates were supplied by A. A. Duncan, Department of Agriculture, Wanganui.)

Table II.—ESTIMATES OF NATURAL AND ARTIFICIAL PONDS

No. of birds Approximate area Order of Type of flock counted in flock birds feeding on density/acre Post-moulting 966 2.0 acres 485 429 1.5 acres 285 393 5.0 acres 80 293 1.0 acres 295 146 1.0 acres 145 127 1.5 acres 85 85 0.5 acres 170 Non-breeding 121 2.0 acres 60 91 0.5 acres 180 50 1.0 acres 50

Table III.—ORDER OF DENSITY OF BIRDS PER ACRE

No. of birds Direct Distance Approx, area and Direct Distance Approx, area and in flock to water type of water to water type of water 89 100 yards river 100 yards river 93 immed. adjacent 0.5 acre pond immed. adjacent 0.5 acre pond 127 immed. adjacent 0.25 acre pond immed. adjacent 0.25 acre pond 165 immed. adjacent 0.25 acre pond immed. adjacent 0.25 acre pond 227 immed. adjacent 1.0 acre pond immed. adjacent 1.0 acre pond 554 immed. adjacent 3.5 acre pond immed. adjacent 3.5 acre pond

Table IV—DISTANCE OF FLOCKS TO WATER

100yds 100ydsBeyond Total No. Distance of Pond from Trees and less 400yds 400yds of Ponds Ponds with broods 28 17 6 51 55% 33% 12% Ponds without broods visited regularly 17 36 40 93 18% 39% 43%

Table V.—DISTANCE OF PONDS FROM NEAREST TREES