has occurred because feeding specialisation brings the greatest advantages and has been favoured by natural selection. Partitioning the habitat would necessitate birds moving from one suitable micro-habitat (say a species of tree) to another, and it would depend on the pattern of the total habitat how much time would be wasted in the process. But adaptation to a comparatively broad habitat structure, for instance, to arboreal or ground feeding, must in turn impose physical limitations which restrict the diversity of feeding adaptations; in practice, bill size and shape is about all that can be much modified to suit the collection of different foods.
From the viewpoint of zoo-geography, the Palaearctic has existed as an entirety for sufficient time to ensure that most niches are filled by highly efficient species. Furthermore, man and birds have lived side by side since Neolithic times, so that the new habitats created by agriculture and man’s other activities have been occupied by the species best suited to them. The same is not true of Australia and New Zealand, which were cut off from the main centres of evolution at an earlier stage, one result being that primitive marsupial mammals were not replaced by the better adapted placental mammals. Birds are less insular, however, and the native avifauna of Australia seems to be the best fitted to occupy the niches available. Thus of at least 24 bird species deliberately introduced into Australia in the past, only 12 have become established. It is significant that only the blackbird has managed to invade native forest, the remainder existing in areas of recent agricultural development or urbanisation. But introduced birds like the feral pigeon, starling and house sparrow are better equipped to occupy the man-made niches than the native fauna, simply because these are species which have already been selected to occupy a man-made environment. New Zealand has an impoverished avifauna compared with Australia on which it has depended for colonisation, and the process is still incomplete. In consequence, fewer niches are saturated in New Zealand, so more exotic species have been successful. Of 130 species originally introduced, 24 have become established, although apart from the blackbird, chaffinch and redpoll which appear to be filling unexploited niches, most are again restricted to man-made habitats. Hawaii, which is even more isolated, and not saturated by a wide diversity of species, must have even more vacant niches, for, according to Elton (1958), of 94 birds introduced 53 have become established, some deep in the native forest.
According to Middleton, the European goldfinch has been successful in Australia and New Zealand only in the man-made agricultural areas, to which none of the native Australian Ploceid finches were adapted. In contrast, the European goldfinch has not been a successful bird in North America because it has virtually the same ecological requirements as the fitter endemic American goldfinch. Only one small colony of European goldfinches became established near New York, though these have since vanished when their habitat was destroyed for building purposes. Again the European house-sparrow has been highly successful in Australia, over roughly the same range as the goldfinch, whereas the introduced greenfinch is more restricted as it has rather more conservative ecotone requirements. It is interesting that this reflects a trend occurring today in Britain; the greenfinch is declining with the loss of hedgerows and woodland edges, while the goldfinch and linnet are increasing.
To return to New Zealand, it is noticeable that the birds which have become pests in agricultural areas, apart from being introduced species as one would expect from the comments above, present the same kinds of problems as they do in Britain. I am grateful to Dr P. C. Bull for allowing me to give details. As we shall see, skylarks (see here) are locally troublesome in Britain to young seedling crops such as lettuce. Near Hastings, N.Z., they and house sparrows have together been responsible for damaging asparagus and other seedlings. Both blackbird and song thrush and also the starling, resort to orchards in the dry season after breeding and cause considerable damage to all kinds of fruit, ripening pears, cherries and grapes. Redpolls do considerable damage to apricot blossom in their search for insects, and blossom searching is a habit which is increasing in Britain (see here). Locally in Britain, linnets peck out the seeds from strawberries (see here), while in N.Z. goldfinches do the same.
Various attempts were made to introduce the rook into N.Z. from 1762 onwards but only 35, liberated near Christchurch in 1873, seem to have thrived. The species occurs in five localities on the yellow-grey earths in the east of the county, generally where cereal growing occurs. Rooks at first increased very slowly but there was a rapid increase between 1935 and 1950, and a final levelling off with the density of birds in their favoured areas becoming virtually the same as that in Britain (around 16 nests per square mile). At Christ-church, the population increased from 1,000 birds in one rookery in 1925 to 7–10,000 in 1947 (13 rookeries), since when the numbers have remained roughly constant with 19 rookeries in use. Until 1926, the rookeries were in eucalypts, probably the favourite tree, but following a disease epidemic which killed these trees the birds changed to pines. Bull points out that the rate of increase of rooks has been slower than that of other introduced passerines and attributes this to their gregarious nesting habits and their need for group stimulation, and to an early shortage of suitable habitats. A feature of N.Z. rookeries is their very large size compared with British ones (rookeries of over 1,000 nests are quite common), and the traditional return to the same nesting sites may partly explain the slowness of expansion. (The large size of rookeries, and difficulties in getting sufficient food locally may also explain why the birds seem to lay, on average, smaller clutches in N.Z.; 3.4 eggs against 4 + in Britain, (see here), though more data are needed to establish the point.) Frequently, only when man actively disturbed these large rookeries did they become fragmented in surrounding areas, often with a rapid increase in total bird numbers in the district. As in Britain, rooks uproot seedling peas and corn, take ripening peas (and pumpkins) and maize and are also partial to walnuts.
The number of closely related birds which can live in the same habitat without competing for food depends to a large measure on the degree of stability within the environment. Marked fluctuations occur on English farmland, not only because of the changing seasons, but also because ploughing, harvesting and other farm operations impose drastic changes. As a result, the farmland birds occupying the various niches available for ground-feeders show a wide character displacement; we find a plover, three passerines (rook, starling, lark), a partridge and a pigeon, other species being only transient visitors, or primarily dependent on other habitats. No bird can afford to be too conservative in its niche requirements in a fluctuating environment, while the need for each species to show more tolerance reduces the number of ecologically isolated forms. Therefore, we should expect modern farm mechanisation, which enables whole farms to be ploughed within a fortnight, to be detrimental to bird life compared with the old methods which ensured some degree of stability by leaving land fallow and by transforming stubbles into bare ground more gradually. Klopfer and MacArthur (1960) have similarly emphasised that the major factor accounting for a decrease in the number of species away from the tropics, while the number of individuals of each species increases, does not result from a decrease in habitat complexity, but to a decrease in the similarity of coexisting species. The principle can obviously be extended to any situation where man simplifies the environment.
An important feature of complex ecological communities is that interactions between members damp out oscillations in the numbers of any one species (see here) and so help to introduce a high degree of stability and energy utilisation. For one thing, available food is more fully exploited, which is not the case in arctic environments for instance, where considerable seasonal changes occur. Hence, the amount of energy needed to maintain a stable community is less than that required for an unstable one. Man’s activities have tended to reduce complexity and introduce monotony, through monocultures of crops, uniform stands of trees, or rows of similar houses. In consequence, the animals inhabiting these environments usually fluctuate much more than those of more complex ecosystems, often to the extent of becoming pests (see here). One feature of stabilisation is that natural selection can favour anticipatory functions – for example, the breeding season of northern birds has become approximately geared to seasonal daylight changes – in unstable environments opportunism must set more of a premium. Because