BIRD POPULATION During 1899-19o1 F. L. Burns carried out a sectional bird census at Berwyn, Chester County (Pa., U.S.A.), and reported 588 pairs of native birds and io6 of English sparrows (Passer domesticus) on one square mile of ground. Since that time a certain amount of work has been done on both sides of the At lantic, but data are still unsatisfactory for purposes of generaliza tion. The available British results are fairly accurate and com plete but very few, while the American are more copious but less reliable. In the summer of 1907 S. A. Forbes found an average of 600 native birds per square mile in southern Illinois, which was less than half that obtained by the later national census. He esti mated for the whole State 114 English sparrows per square mile, but a large number of these were young. In 1914 F. L. Burns re peated his census at Berwyn and got 617 pairs of native birds on the area which had yielded 588 at the opening of the century— a density still below one pair to the acre.
In the same year the national census obtained data on a large scale for the north-eastern States and much more sparsely for the remainder. Observers were not always competent, and they sometimes misunderstood in structions. Nevertheless it was considered possible to work out a preliminary average for the States north of North Carolina and east of Kansas, the figure arrived at being 1 i 9 pairs per I ooac. for farms and 175 per i ooac. for woodland. In 1915 the average for farms rose to 125 pairs, and the average for woodland to
The Plains region from Oklahoma to North Dakota reported 125 pairs per I ooac. ; the average of all counts from the southern States was 76 pairs of 23 species on 58ac. ; and in all States east of the tooth meridian density of farms was roughly 125-131 pairs per 1 oo acres. While the averages were therefore in reasonably close agreement the separate areas varied greatly; 4oac. of native prairie in South Dakota supported only 12 pairs, and 8oac. in Kansas only 31 pairs (or 4o per Iooac.) while a 4oac. farm with orchards adjoining the latter had 49 pairs of the same species, yielding a density over three times as high. Virgin land in Colo rado gave a figure as low as five pairs per 1 oo acres. Toward the other extreme a 5 2ac. peach orchard in Ohio held 1 so pairs; a 4oac. farm in Connecticut 188 pairs (including eight of spar rows and 15 of starlings) while under strict protection eight ac. at Olney, Ill., were made to hold 7o pairs of native birds, and five ac. near Washington (D.C.) possessed 135 pairs of 24 species (including a colony of 74 pairs of purple martins). An 8oac. tract at Albany (Mo.) held in 1914 298 pairs, and 23ac. of built-up village area at Chevy Chase (Md.) showed 148 pairs of native birds and 13 of English sparrows, the most numerous being the American robin with 19 pairs. The 256ac. campus of Cornell university at Ithaca (N.Y.) gave sectional densities ranging from I o4 to 412 pairs per 1 ooac. with an average of 225, 56 of these being house-sparrows. An average of all reports for two years gave seven pairs of American robins and five and one half of English house-sparrows over all farms covered; Burns estimated the number of house-sparrows in the United States east of the Mississippi at 16 5,000,000.
It will be clear from what has been said about territory that the reliability of these figures is seriously compromised by the fact of their being expressed in pairs instead of individuals. To as sume, as was here done, that a singing male represents a pair is to beg an important question; there is no doubt that the males of many species reduce their volume of song, or actually discon tinue it, at a fairly early stage. A large proportion of song de livered at the height of the breeding season is the product of un mated males. Such a census implies, moreover, that all birds secure mates every season, which is far from being the case. The same criticisms apply to some of the European data, next to be considered.
R. J. Ussher and R. Warren (Birds of Ireland, 'goo) made an incomplete but interesting at tempt at an inventory of Irish heronries, which suggests that the number of breeding herons per I o,000ac. in Ireland at the end of the 19th century was between three and nine, so that the total population might be estimated at about 20,000 birds. H. Boyd Watt, in a series of papers (Ann. Scot. Nat. Hist.,
put the number of heronries in Scotland including four or more nests at about 19o. Reference should also be made to the full census of British heronries undertaken during 1928 (British Birds, vol.
C. J. and H. G. Alexander found, in 1908, 555 pairs of migra tory birds breeding on eight sq.m. at Tunbridge Wells, and 26o pairs on four sq.m. at Wye, Kent. In the first case the willow warbler, Phylloscopus t. trochilus (188), whitethroat, Sylvia c. communis (Io1), chiffchaff, P. c. collybita (56), and tree-pipit, Anthus t. trivialis (S5), were considerably the largest items; in the second the willow-warbler (88), tree-pipit (35), whitethroat (30) and nightingale, Luscinia in. megarhyncha (25). E. M. Nicholson on a 4oac. estate near Haslemere, Surrey, obtained in 1926 a summer population of about 290 adult birds, all those which bred upon or habitually foraged over the area being in cluded. They reared altogether about 453 Young, the proportion of fledglings varying considerably with the species. The main items were: chaffinch, Fringilla c. coelebs (35 adults and 49 young) ; British robin (3o a. and 45 y.) ; willow-warbler (29 a. and 35 y.) ; blackbird, Turdus m. merula (26 a. and 35 y.) ; star ling (16 a. and 37 y.) ; song-thrush, Turdus philomelus clarkei (14 a. and 21 y.), and hedge-sparrow, Prunella modularis occi dentalis
a. and 27 y.). A third of the breeding stock consisted of summer migrants, but the winter (December) population of this same area was hardly changed at 258, of which finches num bered 8o (two per ac.; chaffinch and lesser redpoll, Carduelis linaria cabaret, each 32) ; titmice 74, woodpigeon, Columba p. palumbus, 21, and starling 13. B. D. Nicholson by a monthly census of 2oac. of private grounds at Alloway near Ayr, Scot land, obtained on Jan. 23, 1927, 136 birds; on Feb. 20, 262; on March 20, 213 ; on May 1, 236, on May 29, 23o, and on June 29, 209. In January the dominant species were the blackbird (3o), chaffinch (28) and blue titmouse, Parus coeruleus obscurus (24). The sharp rise during the next month was accounted for mainly by an increase of 68 chaffinches; on May 1 the chaffinch still showed 47, and on May 29, 46, or more than two per acre. The rise of the throstle from o in Jan. and eight in Feb. to 36 in May, and of the greenfinch, Chloris c. chloris, from o to 24 in the same period, together with the decline of the blue titmouse (Jan.,
Feb., 45; May, 5; June, o), were the most striking fluctuations. The general density of about
birds per acre may be taken as a typical figure for such compact and highly favourable areas; adding the next in order of numbers (starling and British robin) the dominant species are equally characteristic. In Kensington Gardens, London (an area of 275ac., where feeding on a large scale has produced a highly artificial bird community) far higher winter densities are recorded, and the average for the whole area in November worked out at 14.47 birds per acre.
On agricultural land the population, as a rule, is much more sparsely distributed. The Oxford bird census, by repeated counts during 1927-28, obtained winter densities of one–two birds per acre and a summer total of 444 on 231.6ac., the dominant forms being blackbird (75), chaffinch (58), throstle (36), whitethroat (24), and British robin (22). But on fallow in autumn, concen tration of finches, skylarks, etc., produced densities over limited areas of over 3o per acre. On rough grass-land extending over 153ac. of the South Downs near Seaford, Sussex (with some low scrub, chiefly gorse), J. F. Thomas carried out a bird census 20 times during Oct. and Nov. 1927, the density fluctuating between 0.1 1 and 4.44 per ac. with an average of 1.24. The detailed re sults showed that, at any rate in certain localities, the numbers and proportions arrived at for the various species might vary widely according to the day on which the census was taken. Only those counts which are confirmed by several checks can be accepted as satisfactory.
More data are available in the case of species shot for game or protected in sanc tuaries. It is estimated that the number of breeding pairs of her ring gulls (Larus argentatus) on the German North sea coast rose from 5,000 in 1906 to nearly 18,000 in 1923. The rise seems to have been general and is not attributable to the establishment of state sanctuaries, which harboured only 7,000 of the 1923 stock. Nevertheless, the multiplication of gulls at the expense of other species was in certain cases very marked; on Memmert the 1906 population of 55o breeding pairs included 8o of herring gulls and 400 of various species of tern, while the 1923 population of 9,700 pairs included 5,50o of herring gulls and 4,020 of terns. These figures were obtained with a high degree of accuracy by marking the nests with numbered pegs; a system which had to be discontinued elsewhere on account of the disturbance involved.
In the Alps of Dauphine (south-east France) a sectional census by E. M. Nicholson during July–Aug. 1927 resulted in densities varying from approximately 0.4 birds per i ooac. in barren valleys between 7,24oft. and 1o,5ooft., to 7.8 in the zone between 5,300
feet. The general distribution was two to four birds per acre between 2,3 7of t. and 5,000f t. ; about six per acre for 5,00o 6,000ft. and about o.o5 per acre above 6,000ft.
From all available figures it appears that the bird populations of Europe and North America are remarkably similar, having a general average over large areas of one–four birds per acre, with a handful of species dominant. Satisfactory data from other parts of the world are wanting. Bird population tends to be lowest on wide tracts with sparse cover, and highest in small pockets formed by orchards and homesteads, open towns or villages, lakes, sea-cliffs, and other obvious points of concentration. In such extreme cases as rookeries of the Adelie penguin (Pygoscelis adeliae) a density above 25o per acre may be maintained over as much as 3o sq.m., but an upward limit is invariably set by the quantity of accessible food resources.
Other factors besides food-supply may limit or frustrate repro duction. A. L. V. Manniche (Meddelelser om Grenland, xlv., 1910) found that in north-east Greenland during the highly un favourable summer of 1907 the entire sea-bird population mi grated early without having been able to breed. The king-eiders, (Somateria spectabilis), paired and took. up territory, "but no signs of open sea water appearing as yet the birds assembled in small flocks instead of building their nests, roving about for some days in the fresh waters and leaving the country at last altogether." In all regions bad weather at a critical period may destroy a large part of the product of the breeding season, and beside such catastrophies there is abundant evidence that fertility and mortality among young birds, before and of ter hatching, vary considerably according to locality and season. Clearly in the case of a more or less stationary population the number of eggs laid per annum must approximately equal the number of deaths from all causes. Among nidicolous young the losses at an early stage are very heavy. Thus B. D. Nicholson on the 2oac. at Allo-, way already described, observed, during 1927, 156 nests containing 687 eggs (disregarding those in which no laying took place). Of these only 420 were hatched and 30o fledged. Other British statistics, which are very few, show between 4o% and 6o% of eggs laid as being destroyed before fledging, and a similar rate of loss is probably maintained until the young are able to fend for themselves. In an average year, apart from migration, the bird population at the end of the breeding season is probably only about double what it was at the beginning. In tropical countries there is no definite breeding season, nests of many species being met with in every month of the year. J. P. Burkitt, in the case of the British robin, shows reason to believe that not more than o.8 young per pair survive till the following breeding season.
Most small birds breed at the end of their first year, and appear to have an average life in a wild state of be tween two and six years. Marked swallows (Hirundo rustica) have on more than one occasion been recovered at nine years old; gulls, plovers, ducks, herons and other larger birds frequently exceed this span. A certain great skua (Catharacta s. skua), at Herma Ness sanctuary, Scotland, is said to have lived there at least 32 years, and birds-of-prey are undoubtedly capable of still longer life. There is a satisfactory record of an eagle owl (Bubo b. bubo), still living in captivity at 68. But with the great ma jority of species generations are short and turnover rapid. The sex ratio is near equality, except in the case of polygamous species (various gamebirds, etc.), and a few like the European cuckoo (Cuculus canorus), which are polyandrous or promiscuous. Guthrie-Smith (Bird Life on Island and Shore, 1925), has shown that the dark southern skua or Sea-hawk (Catharacta ronnbergi), breeding on islands off New Zealand, frequently if not habitually forms households consisting of two males and one female.
Eliot Howard, British Warblers (19o7-14; Bibliography.-H. Eliot Howard, British Warblers (19o7-14; monograph-detailed observations on territory), and Territory in Bird Life (192o; promulgation of the territory theory) ; United States Dept. Agric. Bulletins No. 187 (1915) and 396 0916; United States Bird Census Reports) ; J. A. Thomson, The Biology of Birds (1923; bibl.; popular digest) ; H. Weigold, Die Vogelfreistiitten der Deutschen Nordsee (1924; German sanctuary statistics) ; E. M. Nicholson, How Birds Live (1927; bibl.; popular account, controversial) ; C. S. Elton, Animal Ecology (1927; bibl.; inter-relations with other animals).
(E. M. N.)
