FURTHER NOTES ON THE TUFTED IN ST. JAMES'S PARK, LONDON

By E. H. GILLHAM

INTRODUCTION THIS PAPER on Tufted Duck (Ay thy a fuligula) in St. James's Park deals primarily with observations for 1956 and 1957. In addition, however, certain information gained during the previous three years, 1953-55, is summarized for the first time with the comparable details for 1956-57, and the general picture of events outlined in a previous paper (Gillham, 1957) is brought up to date. As in previous years, observations were made between late April and early October as follows : — On 90 days (twice in a day on 24) in the 140-day period 18th May-sth October 1956. On 128 days (twice in a day on 39) in the 171-day period 23rd April-ioth October 1957. During both breeding-seasons visits were regular, averaging on 4 days out of 7 in 1956 and on 5 days out of 7 in 1957. Daily counts, and the close scrutiny of at the various feeding points, as outlined previously (Gillham, op. cit.), were continued throughout both years. For almost the entire 1957 breeding-season, the construction of two widely-spaced parallel dams, with a drained pit between for bridge-building, divided' the lake into two separate waters. This was of great assistance to counting and observation for, apart from the movement or attempted movement of several newly-hatched Tufted broods from one portion to the other, there was no other passage of families or of independent ducklings between the Horse Guards end and Palace end lakes. Correction to a previous paper. Under Desertion of young by the female (antea, vol. L, p. 4), the nine broods in 1955 were hatched between 2nd July (not 29th July) and 12th August. Number of pairs. The number of pairs present in the two most recent years was 14 in 1956 and 18 in 1957. Between 1954 and 1957, inclusive, the number of pairs attempting breeding each season ranged from 12 to 18. Non-breeding birds. In 1956, unmated non-breeding birds present during the first half of the nesting-period totalled 6 or 7 adult males and 2 or 3 adult females, and in 1957, 6 to 9 males and up to 3 females. During 1954, another year of intensive watching, there was also a marked surplus over the breeding pairs, likewise mainly males.

413 414 BRITISH BIRDS [VOL. LI

NESTING; HATCHING AND FLEDGING OF YOUNG Span of the nesting period. The nesting period (Fig. i) is based partly on calculations made on 50 broods and partly on field observations on the territorial behaviour of pairs. Observations on behaviour have indicated that the hatching-period should commence from the beginning of June, but these early nesting attempts have so far been unsuccess­ ful. In any case, the number of broods which should have appeared in June is very small, since between the last few days of April and the third week of May territorial pairs were considerably fewer than those seen in the period 22nd May-i3th June.

FIG. 1—To SHOW THE TOTAL SPANS OF THE PERIODS OF NESTING, HATCHING, FIRST FLIGHTS OF YOUNG, AND WING-MOULTS OF ADULT FEMALES, OF TUFTED DUCK ( fuligula) IN ST. JAMES'S PARK, LONDON The nesting season begins with a thin population of breeding pairs, then rises fairly sharply to a peak period during which most birds start nesting, after which there is a gradual decline until the last broods hatch. The flightless period of adult females begins with a small number of flightless birds, rises sharply to a peak period when most females are flightless, and similarly declines gradually. Span of the hatching period. The span of the hatching-period in relation to the spans of nesting and fledging of young is shown in Fig. 1. In some cases hatching dates had to be estimated because the exact age of ducklings was unknown, and it is convenient to group the hatching of broods in 8-day periods: — 27th June-4th July 2 broods hatched 5th July-i2th July 10 broods hatched 13th july-20th July 12 broods hatched 21st July-28th July 15 broods hatched 29th July-sth August 5 broods hatched 6th August-i3th August 4 broods hatched 14th August-2ist August 2 broods hatched (this includes an estimated hatching date for a female incubating on 16th August 1954) VOL. LI] TUFTED DUCK IN ST. JAMES'S PARK 415

Fledging period of individual ducklings. In The Handbook it is stated that young Tufted are able to fly at six weeks. Observations on certain broods in 1953 and 1955 indicated that a fledging period of six weeks could be regarded only as a minimum. In 1956 and 1957 closer observation on broods, especially on nine broods watched until deserted by their mothers at ages ranging from 39 to 43 days, showed that at 42 days old none was capable of flight. One was seen to fly a few yards just above the water at 44 days; from age calculations at least twelve birds could fly well between 7 and 8 weeks old. MIXED BROODS OF TUFTED DUCK AND In 1957, four newly-hatched Tufted ducklings formed part of a similarly aged brood of Mallard (Anas platyrhynchos) and received the same attention from the female as the downy Mallard. This mixed brood was almost certainly the result of a female Tufted's laying in a Mallard's nest, for it is very unusual for female Mallard to tolerate a close association of strange ducklings even of their own species (i.e., ducklings not hatched out in a Mallard's own nest), with their own broods. Both of the two surviving Tufted of this mixed brood eventually deserted their foster-mother when about two-and-a-half weeks old and reared themselves. Laying in the nest of another species is probably by no means rare in the Park, for I have known two newly-hatched Mallard to be part of a newly-hatched Tufted brood and a Red-crested Pochard (Netta rufina) duckling amongst a fresh brood of Mallard. In all cases these ducklings were accepted by their foster-mothers. (Since the completion of this paper there have been three further instances of newly-hatched Tufted ducklings with foster-mothers of another species:— (1) a female Pochard (Aythya ferina) with a brood of six, including three Tufted, on 30th June 1958; (2) a female Mallard with five, including two Tufted, on 8th July 1958; and (3) a female Mallard with four ducklings—all Tufted—on 14th July 1958. The last record, in particular, raises the question whether other species sometimes take over a Tufted Duck's nest in the laying period.) CLUTCH SIZES AND BROOD SIZES Brood sizes are based solely on the number of young in the various broods of tiny ducklings when they were first seen on water. The following summary refers only to 1956 and 1957, the years for which figures are particularly accurate: — Number of young in broods Brood sizes 1 2 3 4 s 6 7 8 9 10 Number of broods of the brood sizes 2 1 1 S 7 3 1 4 1 2 = 27 broods detailed above 416 BRITISH BIRDS [VOL. LI

The average brood size was between 5 and 6 young-, whereas data relating to eighteen full clutches found in 1954 and 1955 (W. G. Teagle, in litt.) give an average clutch size of 9 eggs. There is a possibility that a proportion of eggs fail to hatch as has been reported for the Red-breasted Merganser (Mergus serrator): in a study of twenty-two clutches of this species on Schliemunde, Germany, Pflugbeil (Hamerstrom, 1957) found that of at least 262 eggs only 53% hatched, while the average number of unhatched eggs was 1 to 4 per nest. In St. James's Park there is positive evidence of a few eggs in nests after hatching, for in 1954 Mr. W. H. Punter (Teagle, in Hit.) found that, out of two Tufted Ducks' nests holding 20 eggs, a fifth failed to hatch. Rather less conclusive are my own observations of two different females leaving their newly-hatched young on water and return­ ing to, and sitting on, their nests. This apparently unusual behaviour may have been associated with unhatched eggs remain­ ing in these nests though, regrettably, in neither case was it possible for me to confirm this by inspection. Although in the five years personal opportunities for studying nests have been few, I know of single instances of (a) 2 dead young being found in one nest, (b) a broken egg in a nest, and (c) an egg lying just outside a nest In addition, Mr. Teagle informed me that 4 of 12 eggs—the combined total laid in two nests— disappeared in the incubation period. These incidents point to a combination of factors being responsible for the low average brood size. In considering the average brood size it is necessary to make some allowance for a percentage of broods resulting from second nest attempts. This should lower the average brood size, for it is highly probable that with Tufted and other species the clutch sizes of second attempts are generally 1 smaller than those of first layings—as has been reported for some North American wild ducks (Farrington, 1947). (The following evidence, obtained since the completion of this paper, suggests that a return to the nest by the female after most eggs have hatched is associated with unhatched eggs remain­ ing. On nth July 1958, I was informed by Mr. A. May that a female Tufted Duck, without any young, was seen at a nest containing one egg. I examined this nest the same evening and found, in addition to one cold egg, the usual chips of shells indicating hatching. Originally this clutch consisted of seven eggs —including one lying eighteen inches outside the nest—and on the morning of nth July there was a fresh brood of five, with parent, on the lake.)

AMALGAMATION OF BROODS Amalgamation of two or more broods of tiny young of the same species under one mother is well known with the Shelduck (Tadorna tadorna) and with many kinds of diving ducks. The VOL. LI] TUFTED DUCK IN ST. JAMES'S PARK 417 single instance of a large grouping of small Tufted in St. James's Park was in 1957 when the care of three newly-hatched broods (Nos. 16, 17 and 18 in Table II) devolved on one of the three females within about sixty hours of these broods' reaching the water (see Table II). In his study of the uniting of broods of the Red-breasted Merganser, Bergman (1956) found that, when two females with their broods happen to meet, the parents react more or less aggressively, the weaker mother often leaving its young for some minutes. If the aggressive or escape behaviour is strongly activated, this behaviour releases a marked increase in the uniting drive of the young which give their distress calls and huddle together. If the broods are of the same age and the uniting drive is strongly activated, they join each other, too, and swim away conducted by the stronger female. This appears to be a sound explanation for some amalgamations, though I would add to it that I consider that the weaker mothers may be those whose wing-moult is soon to begin. From Bergman's explanation, however, it is not to be implied that, say, the large broods of 30, 40 and 80 Shelduck of various sizes (c/. Gillham and Homes, 1950), or the group of 85 two-weeks-old White-winged Scoters (Melanitta fusca deglandi) and the broods of 20 or more (Aythya affinis) mentioned by Hochbaum (1944), result solely from "battles" between brood females when they meet. Hochbaum points out that the instinct to follow is strong amongst deserted flightless ducklings at Delta (Manitoba, Canada) and the same is particularly true of Shelduck on the North Kent foreshores. Like Mallard, some Tufted mothers attack stray ducklings of their own species which try to join their broods, and this may in part explain why the uniting of Tufted broods has been infrequent. From close observation on the 1957 amalgamated brood, it was often noticed that the foster-mother attacked individuals of her large brood, which pointed to her recognizing that some of these ducklings were not of her original family. Amalgamation of a few ducklings with a brood of similar age probably occurs with fair regularity in Tufted Ducks, especially when the brood female's hostile behaviour to strange young with her family is beginning to wane. Fluctuating daily brood totals in Tables I and II indicate that temporary reductions and temporary increases in the sizes of a family frequently occur. UNATTACHED YOUNG In 1955, 1956 and 1957 a few lost or deserted ducklings were present on the lake when only a day or two old. It was not uncommon for a female to lose one or two of her brood within a few days of getting them to water for the first time. Sometimes these lost ducklings rejoined the brood, but sometimes they remained alone and reared themselves. In cold and wet weather they died quickly, but if good weather predominated during their TABLE I—DAILY STATE OF EACH BROOD OF TUFTED DUCK (Aytkya fuligula) IN ST. JAMES'S PARK, LONDON, IN 1956 NOTES ON TABLES I AND II: (i) Brood identification in the. field. This was to some extent simplified by noting permanent or transitional plumage features in some of the brood females. The differ­ ing brood sizes and the use of special resting places by some families for at least a few days were also guides, and in 1956 the number of broods was very small.- (ii) Identity number of brood, and brood female. Broods which reached water in the care of a female are numbered from 1 to 37: these numbers are used in the text. (iii) Daily brood totals. When counts of broods were made twice in a day the number of young in the family had sometimes altered. The temporary tagging on of lost or deserted ducklings and the temporary absence of a mother for a wing-exercise flight, or to feed elsewhere, contributed to fluctuating brood totals. The most satis­ factory count for the day has been entered. (iv) Broken line over daily brood totals. The length of the broken line shows the overlap between the female's brood duties and her wing moult. (v) Ages of broods. The actual ages of broods or of some young in a particular brood, where mentioned in the Tables or in the text, may be up to a day more than stated, the extra time representing the period between hatching and reaching water. (vi) A sterisk over certain brood totals in Table I. This indicates that on the dates starred one or two young in those broods were not of the original family. Most of these strange young were markedly different in size from the rest of the brood. This would also apply to odd young in some broods in Table II, but no special indication is given there because of the obvious examples of increases (e.g., the combining of broods Nos. 16, 17 and 18 ; and the increase in the size of brood No. ai). (vii) Special note 'A' relating to brood No. 3. In the paragraph on Clutch and brood sizes, brood No. 3 is classified as totalling 6 young on reaching water, for the two lost young (see " loose young " column for 10th July) were near-by and almost certainly belonged to this brood. The subsequent increase in the brood total was of duck­ lings of similar size, except for one obvious newcomer, which suggested the family bad reunited—at any rate temporarily. (viiij Fate of broods Nos. 2, 5, 6 and 34. Not known whether the young were deserted early or died under the care of their parent: in most cases'probably the latter, (ix) Broods Nos, z and 6. Dagger^marks in these daily totals denote that either No. 2 or No. 6 was seen. Unless both were seen identification was impossible. TABLE II—DAILY STATE OF EACH BROOD OF TUFTED DUCK (Aythya fuligula) IN ST. JAMES'S PARK, LONDON, IN 1957 420 BRITISH BIRDS [VOL. LI

first 14 days of life the survival rate was high. Even in spells of good weather some nights are extremely cold and one wonders how these tiny ducklings keep warm. In daytime, however, tiny independent ducklings have been seen to leave the water and, for a short while, nestle close to strange Tufted broods resting ashore, so there is no apparent reason why such behaviour should not occur at night, too.

NANNY FEMALES "Nannies" are non-breeding or post-breeding unattached females (either full-winged or flightless) which assume the r61e of mother over small, temporary or permanently lost ducklings. Although most nannying has been observed only for parts of single days the behaviour calls for comment. During 1953, 1955 and 1956 observations were inconclusive, but at times counts indicated either that extra females with odd medium-sized downy young had suddenly appeared when a steady decline in the total of brood females was taking place (through desertions), or that even the most accurate counts of family parties were at fault. In 1957, however, a closer watch was kept for this behaviour and the first conclusive evidence was obtained for Pochard. Through­ out the breeding-season there was only one brood of this species on the Horse Guards Lake and at 07.40 hours on 1st July this family consisted of the female and two surviving young then aged seven days. At 12.00 hours on the same day this bird had only one young, while on a different part of this water another female Pochard was behaving exactly like a mother and attending the second downy duckling. The following day both ducklings were with their own mother in her usual corner of the lake. By the end of the 1957 season several cases of nannying by Tufted females were also established. In one instance a full- winged female (known by a narrow strip of white at the base of her upper mandible throughout the nesting season to have been brood female No. 13 in Table II, who had' lost her brood two weeks beforehand) was seen to mother the only lame duckling of the amalgamated brood (Nos. 16, 17 and 18) on both 23rd and 26th July. On the first date this female was seen to brood the lame youngster ashore and on the second occasion she drove off two male Mallard swimming close to the same duckling. Another female started to look after a three-weeks-old duckling just after shedding her flight feathers and continued to attend the same one for the next fourteen days. In every case the nannies' behaviour was indistinguishable from that of a typical mother attending her own brood. Although these nanny females do not appear to have any particular biological significance it is useful to know that they occur, for their presence can distort the counts of family parties and perplex the person making the counts if he or she is unaware of their existence. An incident not dissimilar to the cases of nannying already VOL. LI] TUFTED DUCK IN ST. JAMES'S PARK 421

mentioned is given by Hochbaum (op. cit.) who quotes an observation by Monro of a yearling (non-breeding) female Barrow's Goldeneye (Bucephala islandica) which attempted to usurp the brood of an older female. DESERTION OF YOUNG BY THE FEMALE Pattern of desertions. The pattern of desertions was similar to that described for 1953 and 1955- In 1956 all females had abandoned their broods by 1st September, and in 1957 by 13th September: during these two years no mother stayed with her young later than the 43rd day after they had reached water. Desertion of young within a few days of their reaching water. It has been mentioned in an earlier paper (Gillham, op. cit.) that a brood was deserted when a few days old, and it will be seen later here that females may have to desert their young within a fortnight of hatching if they are to proceed to a moult area out­ side the Park. In 1955 and 1957 there were instances of 3 broods being deserted within four days of their reaching water (in one case on the very day they reached water). However, not all early desertions may be correlated with the approach of a brood female's wing-moult. Since two females have been seen to leave their newly-hatched young and return to sit on their nests shortly after taking their ducklings to water for the first time, some early desertions may result through an urge to return to the nest (? associated with unhatched eggs remaining there) overcoming the urge to attend the brood. In the first of these two occurrences the female (No. 7 in Table I) did not lose her brood, although only one of the six ducklings managed to get back to the nest. In the second instance, however, the female (No. 13 in Table II) returned quickly to her brood after sitting on the nest, but an hour or so later the same four young were roaming the lake alone, which suggests the possibility that the female lost them through a further visit to the nest. This brood probably died the same night as they were not seen again. Their mother, whose distinctive recognition mark has already been described, was still present two weeks later at which time she had' not begun her wing-moult. If a nest is in such a position that it is impossible for the young to get back to it, and, more particularly, if a brood female does not call the young to follow when this urge to return to the nest dominates her, it is not difficult to visualize how losses of very small young can occur. It may not be without significance that the nest-sites to which these females returned, after taking their young to the water, were the only ones to be kept under close observation for a long period.

MORTALITY OF YOUNG The average mortality in the 1956 and 1957 fledging-periods 422 BRITISH BIRDS [VOL. LI

was found to be 50% of all ducklings which reached water. It is believed that an average of 70% of all casualties in these two years were ducklings aged between one and seven days. The heavy mortality in 1957 (and in 1954 as well) was almost entirely due to this period of maximum vulnerability coinciding with cold and wet weather. Ducklings of up to about ten days of age feed largely on and matter taken from the surface, and in cold wet weather there is an obvious shortage of the former. An absence of vital food coupled with bad weather in this critical period of their lives must be an important factor affecting survival. In 1957, the first six broods (see Table II) hatched in the cold and wet spell between 8th and 18th July; all died within a week of reaching water. These 29 ducklings represented ca. 53% of the total casualties for the season.

Total mortality of young which reached water

Number to reach Number to reach Percentage of young to die Year water flying stage in fledging period

195° 45 25 44-4% 1957 107 52 5M%

Mortality of young within 7 days of reaching water

Percentage of total Number to die in Number to die in casualties which died Year fledging period their first 7 days in their first 7 days

I9S6 20 ca. 9 ca. 45%

1957 55 ca. 44 ca. 80%

ADULT FEMALES IN WING-MOULT Number of adult females completing their wing-moult on the lake. In 1956, 11 adult females (out of a maximum population of 17) moulted their flight feathers on the lake; and in 1957, 12 (out of 21). Thus, in the two years, 60% of all adult females, mainly breeding birds, stayed to complete their wing-moult, most of them departing soon after regaining their powers of flight. Compared with the period 1953-55, more adult females completed their flight­ less period on the lake. Span of the wing-moult period. Between 1953 and 1957, inclusive, a total of 37 adult females (roughly 45% of all females present in the five nesting-seasons) completed their wing-moult on the lake, most of them being seen in this condition, initially, just after shedding their flight feathers. In view of the intensive observation during the moult period it is unlikely that more than an odd flightless bird was missed. The VOL. LI] TUFTED DUCK IN ST. JAMES'S PARK 423

data on these 37 birds form the basis of the span of the wing-moult period shown in Fig. 1. It is possible that one or two of these females came to the lake to moult from a breeding area outside the Park. Duration of wing-moult in individual birds. From close observation on four birds the duration of the flight­ less period was between 2f and 4 weeks, which agrees with the period given by Hochbaum for some of the Delta diving ducks. Flightless brood females continuing to attend their young. During his studies at Delta, in Manitoba, Canada, Hochbaum found no evidence of female diving ducks moulting their flight- feathers while mothering their broods. In a previous paper (Gillham, op. cit.), however, I gave two instances of such an overlap between part of a mother's wing-moult period and her brood-rearing duties, to which may now be added a further 3 occurrences in 1956 and 8 more in 1957. Although these 13 mothers ultimately deserted their broods before the young could fly, they tended their ducklings throughout part (and in one case all) of their own flightless period. Taking only the 11 females in 1956 and 1957 (see Tables I and II), for which the data are most accurate, the duration of the overlap varied up to 22 days, the average attendance on young being for 14 days following the beginning of a brood female's wing-moult. The ages of the eleven broods at the time when their mothers shed their flight- feathers were 14, 15, 18, 20, 21, 21, 22, 28, 28, 34 and 38 days respectively. These figures give an idea of the differing ages of broods when their mothers actually became flightless and cause interesting speculation as to what would have been the ages of these eleven broods at the time of desertion had they been left on dates before their mothers became flightless. Does a mother who deserts her brood before her wing-moult leave the young a few days or a week or two before her flight-feathers are ready to drop out? Those female Tufted which leave the lake for a moult area outside the Park probably need to desert their young at least a week before their wing-moult begins if they are to undertake a long flight. Taking as an example from these eleven broods the ducklings which were 14 days old when their mother became flightless, it can be realized how downy young may be deserted' at the age of only a few days through their mother's departure to moult elsewhere.

MOVEMENTS OF BREEDING FEMALES In 1956, between 1st and 10th September, the number of adult females was 4 or 5 (under one-third of the breeding season popula­ tions of adult females), though numbers rose later through the incidence of passage birds. The following year adult females 424 BRITISH BIRDS [VOL. LI totalled 8 from 19th September and 4 or 5 (approximately one- fifth of the breeding season population) from 4th October. In both years the departure pattern—before or after the wing-moult —was similar to previous years except that more females delayed leaving until after their flightless period. The span of the departure period is given in Fig. 2.

FIG. 2—To SHOW THE APPROXIMATE SPANS OF THE DEPARTURE PERIODS OF THE BULK OF THE BREEDING ADULT TUFTED DuCK (Aythya fuligula) AND THEIR YOUNG

FROM ST. JAMES'S PARK, LONDON

ADULT DRAKES The movement of adult drakes in 1956 and 1957 was similar to that in the two previous years. In 1956 most males had departed by 22nd July and on the 31st only two remained, both subsequently completing their flightless period on the lake. The following year most had gone by 27th July, but the 6 or 7 remaining through­ out August had increased to 8-10 during 7th-i6th September, after which numbers fell to 2 or 3 from 19th September. In this latter year 5 adult drakes certainly shed their flight-feathers on the lake and a sixth probably did so. The total number of males completing their wing-moult on the lake was ca. 10 during the five years I953"S7- The earliest that any of these ten drakes first became flightless was the middle of July and the latest time for a bird to regain its powers of flight was about the middle of September. The span of the departure period' of the males which proceed to a moult area outside the Park is shown in Fig. 2.

DEPARTURE OF YOUNG The departure of fully-winged juveniles of the year was similar to that recorded for 1953 and 1955. Counts of the number of juveniles on the lake were as follows: — 1956 31st August, 25 (total to reach flying stage); 6th September, 19; 7th September, 30 (a marked influx of passage birds); nth September, 10; 17th September, 17; 5th October, 12. 1957 13th September, 52 (total to reach flying stage); 25th September, 40; 8th October, 2$; 10th October, 14. From these counts it will be seen that in 1956 a minimum of 50% of the original surviving young left the lake between 1st September and 5th October; and in 1957 just over 70% departed between 14th September and 10th October. To summarize the VOL. LI] TUFTED DUCK IN ST. JAMES'S PARK 425 movements of young for the four years 1953 and 1955-1957 inclusive, it can be said that an average of at least 57% of the surviving juveniles departed within four months of hatching and that most of the remainder left at some time between the middle of October and the following spring. NOTES ON FOOD On 3rd July 1957, a female was watched for about thirty minutes removing a broken egg from her nest. Three times she was seen to enter the water and eat pieces of shell, and she finally swallowed whole the two-thirds-grown embryo. On 19th August 1957, two four-weeks-old ducklings were seen eating the leaves of a white poplar (Populus alba), which they plucked from overhanging branches touching the water. On 28th and 30th August 1957, a total of three five- or six- weeks-old ducklings were seen eating off dead fish some 6-8 inches long. The fish were almost certainly Roach (Rutilus rutilus) or Rudd (Scardinius erythrophthalmus), many of which were noted dead or dying on the lake during this month. ACKNOWLEDGEMENTS I am indebted to W. H. Punter and W. G. Teagle, late of the Ministry of Works, and to A. May, the present Bird Keeper, for supplying me with information concerning nests and clutch sizes. SUMMARY 1. The breeding and post-breeding populations of Tufted Duck in St. James's Park, London, were studied between late April and early October in the years 1953-1957 inclusive. 2. The females deserted their broods in the fledging-period, sometimes when the young were a few days old. 3. The adult males departed for the moulting grounds mainly in July. 4. Surplus non-breeding birds, mainly males, totalled between one-fifth and one-quarter of the whole adult population in most nesting-seasons. 5. Slightly under one-half of the adult females stayed to complete their wing-moult, some mothers becoming flightless while still attending their young. The span of the flightless period and its duration in individual adult females are given. 6. The order of departure of the population was the males first, then the females, and finally the young of the year. 7. Non-breeding or post-breeding adult females sometimes acted the r61e of mother to lost or deserted ducklings for very short periods; such birds are termed "nannies". 8. The average yearly mortality of ducklings was about 50% of all those known to have reached water and over half the total casualties were ducklings of up to 7 days of age. Heavy mortality occurred when cold and wet weather coincided with the first week of the duckling's lives. 426 BRITISH BIRDS [VOL. LI

g. Ducklings, deserted or lost when a few days old, reared themselves if good weather prevailed during the first few weeks of their lives. 10. The average size of 27 broods on reaching water in 1956 and 1957 was 5 or 6 young and probable reasons for the small size of family parties are discussed. 11. Occurrences of mixed broods of ducklings involving different species are given. 12. Notes are given on the spans of nesting and of the hatching and fledging of the young; and on the duration of fledging in individual ducklings. 13. Notes are given on food. REFERENCES BERGMAN, G. (1956): "On the uniting of broods of Mergus serrator and Mcrgus merganser" (English summary of paper in Swedish). Fauna och Flora, 1956 (3): 109. GILLHAM, E. H. (1957): "Notes on Tufted Duck in St. James's Park, London". Brit. Birds, L: 2-10. and HOMES, R. C. (1950): The Birds of The North Kent Marshes. London. HAMERSTROM, FRANCES (1957): Review of Pflugbeil (1956). Bird Banding, xxviii: 102. HOCIIBAUM, H. ALBERT (1944): The Canvasback on a Prairie Marsh. Washington. KIP FARRINGTON, S. (1947): The Ducks Came Bach. London. PFLUGBEIL, A. (1956): "Briiten des Mittelsagers auf Schleimiinde 1955". Die Vogelwelt, 77: 44-47.