THE FOOD OF NESTLING PURPLE IN HOLLAND By D. F. OWEN and G. C. PHILLIPS

FEEDING HABITS AND AREAS STUDIED IN late June 1956 we visited two colonies of Purple Herons ( purpurea) in Holland in order to observe feeding and other behaviour for comparison with our studies in England of the Common (A. cinerea). The colonies visited were at the Naardermeer reserve (10 miles S1E, of Amsterdam) and at Nieuwkoop (15 miles S.W. of the Naardermeer). At the Naardermeer we were told that there were over 150 pairs >f Purple Herons nesting in 1956 and at Nieuwkoop at least 200 pairs, where they bred for the first time as recently as 1946. We spent a week watching the adult feeding, and in addition, collected 56 food-samples from the young at the Naardermeer and 17 from those at Nieuwkoop. All food-samples were collected in the last week of June, and thus the results given in this paper refer only to that period. There can be little doubt that the food of the nestlings varies with the time of year and also from year to year. This paper describes what we found in the food-samples and includes a few notes on the behaviour of nestlings and adults. The Naardermeer consists of a series of lakes connected together by waterways and surrounded by extensive reed-beds (mostly sp.) and fen scrub chiefly alder, birch and sallow. There were two groups of nesting Purple Herons, the main one in a Phragmites marsh, and a small one of 30 pairs (the group we studied) in sallow bushes at the edge of an area of fen scrub. All the nests in this latter group were less than 10 feet and most were only 3-5 feet above the ground. At the Naardermeer many adult Purple Herons were seen hunting for food along the numerous narrow canals and in pools surrounded by reeds. Others were seen hunting along the edges of ditches outside the reserve, but our observations did not suggest that the birds regularly travelled more than a few miles from the heronry for food. In many respects Nieuwkoop is similar to the Naardermeer, but it is larger, and much of it is open to the public (unlike the Naardemeer). Here the Purple Herons nest in an alder wood, though considerable numbers also nest in sallows near-by. The alders varied in height from 6 to 20 feet. At Nieuwkoop many of the adult Purple Herons seemed to considerable distances for food, but we do not know to what type of habitat. In addition, many birds were also seen feeding in the canals and among the reeds, just as at the Naardermeer. At this colony a few pairs of Common Herons nest with the Purple Herons, and Little (Ixobrychas minutus) and Bitterns (Botaurus stellaris) are evidently quite common in the area. At the Naardermeer the only 494 VOL. XLIX] FOOD OF PURPLE HERONS 495 other breeding heron is the , but numbers of Common Herons visit the Cormorant (Phalaerocorax carbo) colony where they scavenge regurgitated food. In Holland all our observations of adult Purple Herons suggested that they fed singly, though often several individuals could be seen in quite a small area. They seemed to feed closer together than Common Herons in England, but this might be due to the greater concentration of breeding pairs in the area and the rich feeding-grounds. In the Camargue, one of us (D.F.O.) has watched Purple Herons feeding and here also they were always solitary, quite unlike the Little Egrets (Egretta garsetta) and Night Herons (Nycticorax nycticorax) which are distinctly social.

METHODS OF OBTAINING FOOD SAMPLES AND THE BEHAVIOUR OF NESTLINGS WHEN ALARMED Young Purple Herons, like many other -eating birds, readily regurgitate food when alarmed, and this is a useful means of studying the food of such —see Van Dobben (1952) for the Cormorant, and Owen (1955) for the Common Heron. Probably the birds regurgitate to scare or to divert the attention of their enemies, and, possibly, once relieved of food they are able to escape from the nest more easily. Using this method of obtaining food-samples it is not easy to get food from nestlings less than 10 days old, but some samples were collected from small young by gently massaging the stomach region soon after they had been fed. All the items of prey found in the food-samples were identified on the spot and left in the nest where they were eaten again by the nestlings. Samples were collected from about 50 different nests, some nests being visited more than once. At the Naardermeer all the samples were collected in the early morning and in the evening, at which times there appeared to be a peak of feeding- activity (as in the Common Heron in England). We found that young Purple Herons regurgitated more readily than Common Herons, and rarely attacked the observer, unlike Common Herons which normally attack. There was also a greater tendency to leave the nest and plunge into the undergrowth: even quite small young of about two weeks old would leave the nest when alarmed. Hanzak (1949-50) reported that in Czechoslovakia young Purple Herons 8-10 days old readily left the nest when alarmed and returned later. Related to these differences of behav­ iour between the two species of heron, it might be mentioned that the nest of the Common Heron is considerably enlarged and made more cup-shaped' during the incubation-period and when the young are still small. When the eggs are laid the nest is flat, but by the time they hatch it is distinctly cup-shaped. This seems to serve the function of keeping small young in the nest. We found that the nests of the Purple Heron were much smaller than those of the Common Heron and there was hardly any sign that the nest had been built up during incubation. This is probably related 496 BRITISH BIRDS [VOL. XLIX to the fact that the young Purple Herons leave the nest when disturbed and any building up of the nest might tend to prevent them from doing this as efficiently as they do.

SIZE OF PREY We estimated by eye the length of most of the fish found in samples and found that, like the Common Heron, the Purple Heron catches fish selectively, fish of 110-200 mm. (ca. 4-8 inches) in length being much more numerous than either smaller or larger individuals. The size of fish taken by the Purple Heron in Holland was very similar to that taken by the Common Heron in England (Owen, 1955). This is of particular interest as smaller fish must have been more numerous than larger fish of the same species. All eels taken were larger than 150 mm. (6 inches) and smaller than 400 mm. (16 inches)., suggesting that these fish were also being taken selectively. We could find no evidence of adult Purple Herons bringing smaller prey for the small young. Once, at Nieuwkoop, we found a pike 350 mm. (14 inches) long and' a rudd 240 mm. (gf inches) long in the bottom of a nest containing three young about two weeks old. Obviously these fish were much too large for the young birds to swallow and they had been rejected. Adult Common Herons almost always partly digest a large fish before giving it in pieces to small young, and presumably this normally happens in the Purple Heron. As the prey found in the food samples varied considerably in size

TABLE I—LARGE PREY IN FOOD-SAMPLES PROM NESTLING PURPLE HERONS (Ardea purpurea) IN HOLLAND, JUNE 1956 Naardermeer Nieuwkoop (56 samples) (17 samples)

Pike (Esox lucius) 12 3 Carp (Cyprinus carpio) 1 Gudgeon (Gobio gohio) _ 2 Tench (Tinea tinea) 7 2 Rudd (Scardinius erythrophthalmu.i)* 42 16 Bream (Abramis bra-ma) 9 — Eel (Anguilla anguilla) 9 1 Perch (Perca fluviatilis) 10 13 Ruffe (Acerina cernua) — 2 Burbot (Lota vulgaris) 2 _ Water Vole (Arvicola amphibius) (adults) 14 7 (young) 28 1 Mole (Talpa europaea) 10 1 Shrew (Sorex sp.) — 1 Frog (Rana sp., mostly temporaria)* 8 2 Skylark (Alauda arvensis) (young) 1

TOTAL 'S3 Si

*Note: " Rudd" may also include Roach (Rutilus rutilus), and a few of the frogs may have been Rana esculenta. VOL. XLIX] FOOD OF PURPLE HERONS 497 it would be misleading to compare directly the numbers of each species without making- some allowance for size. This difficulty has been partly overcome by dividing the prey into two categories : Large prey. All eels more than 150 mm. (6 inches) long, all other fish more than 100 mm. (4 inches) long and all and birds. Within this category there was some variation in size, but it is not so great as to give misleading results. Small prey. This includes all fish smaller than the above and all , but very few small prey were found in the food samples collected from the nestling Purple Herons. SPECIES TAKEN Most species of prey found in the food-samples were easily recognisable in the field, the greatest difficulty being experienced with the closely related rudd and roach. All fish in good condition which might have been either of these species were found on close examination to be rudd, and it therefore seems likely that most were of this species. Naardermeer. Table I shows the large prey found in the food-samples collected at the Naardermeer. A total of 153 items of prey were identified, almost entirely fish and mammals. The commonest species were rudd (27%) and water voles (27%). Many young water voles were taken, and often it seemed that a nest of these mammals had been found by the adult Purple Heron, as a sample would contain one adult water vole and several young. Adult Purple Herons were often seen feeding along the edges of ditches and doubtless this would frequently lead to the discovery of a nest. Besides rudd, other species of fish found rather frequently in the samples included pike, eel, bream and perch, all of which would be common in the immediate vicinity of the heronry. Purple Herons were sometimes seen feeding in open fields and this would account for the moles present in the samples. Frogs were numerous in the Naardermeer and probably very available to adult Purple Herons. The single

TABLE II—SMALL PREY IN THE POOD-SAMPLES FROM NESTLING PURPLE HERONS (Ardea purpurea) IN HOLLAND, JUNE 1956 Naardermeer Nieuwkoop (56 samples) (17 samples)

Ten-spined Stickleback (Pygosteu.s pu ngitius) 6 Perch (Perca fluviatilis) 3 Rudd (Scardinius erythrophthalmus) 3 Tench (Tinea tinea) 1 Water larvae (Dytiscus sp.) 27

TOTAL 39 1 498 BRITISH BIRDS [VOL, XLIX found in a sample was an almost fledged Skylark (Alauda arvensis). Possibly a nest had been found by the Purple Heron when hunting for mammals in a field. Table II shows the small prey in the Naardermeer food-samples. We were surprised to find so few insects, especially in view of the fact that other writers give the impression that the Purple Heron takes many insects. A possible reason for this is discussed later. As we have already found in the Common Heron, few small fish of the larger species were taken, and there was a surprising scarcity of such species as sticklebacks. Nieuwkoop. We collected only 17 samples from this colony, all on the same day. Table I shows that the large prey taken were of much the same species as at the Naardermeer. There were fewer water voles and rather more perch, but possibly these differences would not have been apparent had' more samples been collected. The presence of gudgeon perhaps suggests feeding in rather faster flowing water. Except for one small tench, no small prey were found in the samples from Nieuwkoop (Table II).

DISCUSSION The complete absence of and their larvae is of interest, as at both localities these insects were abundant, but as already mentioned, there was a general scarcity of small prey in samples. We have compared our results from Holland with the results of other workers in different countries (Moltoni, 1936, 1948, for Italy; Vasvari, 1930, for Hungary; Madon, 1935, for France). Most of the above workers used stomach contents, not regurgitated food from nestlings. The chief difference (apart from different species of prey due to locality) was that the stomachs contained many more insects than we found in the regurgitated food. This might be due to the fact that more insects were taken, but it also might be due to the slow rate of digestion of remains in a heron's stomach. This means that any analysis of stomach contents gives an over-emphasis on the insect part of the diet. To get over this difficulty only fresh undigested food must be counted and all fragments of insects and fur must be ignored. The three authors quoted above did not make allowance for the different rates of digestion, as we have in the present study. At both the Naardermeer and' at Nieuwkoop many young Purple Herons were found dead in the nest. Often in a brood of four the smallest was dead, and sometimes two out of four young had died. Such mortality strongly suggests food shortage. In some newly-hatched broods all four young were alive, but in every brood more than 14 days old there were only three young or less. We did not see a single brood of five. Work in progress in England on the Common Heron shows that in years when food is readily VOL. XLIX] FOOD OF PURPLE HERONS 499 available broods of four may be 100% successful, but in poor years up to 40,% of the young in a colony may die of starvation, the smallest birds in a brood always dying first. The few facts we have for the Purple Heron in Holland suggest that 1956 was a moderately bad year for the availabilty of food, but possibly other factors were involved, such as cold and heavy storms. The Purple Heron is a summer migrant to Holland and breeds later than the Common Heron. Precise figures are not available, but the difference in breeding-season appears to be 4-6 weeks. Thus, by the time most of the young Purple Herons are half- grown and are needing most food, almost all the young Common Herons will have already left the nest. Certain prey taken by Purple Herons for their young, such as young water voles, would not be available to nestling Common Herons, but possibly a more important difference in the ecology of these two closely related species is that the Purple Heron feeds much more in cover and at small stretches of water, whereas the Common Heron is much more a bird of open country.

ACKNOWLEDGEMENTS We must express our thanks to J. F. Burton and D. F. Dorward for help with the field work in Holland, and also to the Vereeniging tot behoud van N atuurmoniimenten in Nederland and to Mr. O. P. N. Blom for permission to work at the Naardermeer and at Nieuwkoop, respectively.

SUMMARY 1. The feeding habits of adult Purple Herons and the behaviour of nestlings are briefly described. 2. Adult Purple Herons select food between certain size limits for their young, avoiding both very small and very large prey. 3. The food brought to the young at two colonies in Holland in late June 1956 comprised chiefly freshwater fish and mammals. Few insects and other small prey were found. 4. The feeding ecology of the Purple Heron is compared with that of the Common Heron. REFERENCES HANZAK, J. (1949-50): "Report on the breeding of the Purple Heron, Ardea purpurea L., and Great White Heron, Egretta alba (L.) in Bohemia, 1949". Sylvia, 11-12: 85-97. MADON, P. (1935): "Contribution a l'etude du regime des oiseaux aquatiques". Alauda, 7: 177-197. MOLTONI, E. (1936): "Le Garzaie in Italia". Riv. Ital. Orn., (2) 6: 211-269. — (1948): "L'alimentazione degli Ardeidae (Aironi) in Italia". Riv. Ital. Orn., (2) 18: 87-93. OWEN, D. F. (1955): "The food of the Heron Ardea cinerea in the breeding season". Ibis, 97: 276-295. VAN DOBBEN, W. H. (1952): "The food of the Cormorant in the ". Ardea, 40: 1-63. VASVAEI, N. (1930): "Studien iiber die Ernahrung des Purpurreihers (Ardea purpurea L.)". Aquila, 36-37: 267-293.