Following three episodes with (severe) night frost in April 1997 and abundant rainfall in May and June, populations of Vespula germanica and V. vulgaris in The Netherlands collapsed till virtually nil active wasp nests were left in July and (especially) August. Wasp numbers had never been so low since at least 1973, and probably since World War II (Bijlsma 1998). In 1997, 30 egg-laying pairs of Honey Buzzards were found in The Netherlands by members of the Dutch Raptor Group: 7 pairs failed during the egg stage, 7 pairs failed to raise theiryoung, 9 pairs raised a single fledgling and 7 pairs raised 2 fledglings. Local studies showed that in 1997 most pairs failed to lay eggs anyway, i.e. 6 out of 8 in Central Drenthe and 4 out 5 in West-Drenthe. Two pairs were closely monitored to quantify parental care, prey choice, frequency of prey delivery and growth and development of nestlings (see also Bijlsma, van Manen & Ottens 1997). Both pairs were essentially covering the same area of woodland (Boschoord slightly more varied in structure and less dry than Smilde), their nests (in Scots pine and Japanese larch respectively) being 5,6 km apart. Egglaying started on 26 and 25 May in respectively Smilde and Boschoord. Both pairs produced 2 eggs, but only the oldest nestling in each nest survived till fledging; the youngest nestlings died from malnutrition. In both cases, and contrary to normal procedure, parental care at the nest almost ceased two weeks after hatching, both sexes being away from the nest for prolonged periods of time (Table 1). This is thought to have been triggered by food scarcity, forcing both parents to forage continuously. The radio-tagged female of the Smilde-pair was only infrequently recorded near her nest in the second half of the nestling stage, leaving the area permanently a few days before her young fledged (duties entirely taken over by male). The frequency with which fresh prey remains were found during nest visits also sharply declined in the course of the nesting cycle, contrary to expectation based on the increasing food demands of the nestling and experiences gained in earlier years (Figure 1). During July, among prey remains found on the nest wasp combs were frequently encountered, although in smaller numbers than expected. From early August onwards, wasp combs became less common (especially on the nest of the Smilde-pair), apperently being replaced by nests of bumblebees. As the number of larvae in bumblebees’ nests varied between 4 and 8, as compared to an average of 75 larvae in wasp combs (with an average diameter of 55 mm in 1997), this change in food choice is also thought to have been triggered by a steep decline in wasp numbers. During nest visits 115 complete wasp combs were found in 1997, averaging 55 mm in diameter (SD=30.9, range = 35-90 mm). Remarkably, the diameter of wasp combs (i.e. an indication of the size of wasp nests) in 1997 did not increase in the course of the breeding cycle, as is normally the case (Figure 1). Moreover, the average diameter of wasp combs in 1997 (55 mm) was significantly smaller than in 1992, 1993 and 1996 in the same area (mean 79 mm, SD=21.3, range = 24-160 mm, N=227, containing an average of 155-190 larvae per comb)(Figure 2). Moreover, length (and therefore dry weight; see Figure 3) of wasp larvae in 1997 was relatively smaller than in other years. These data again suggest a poor reproductive performance of wasps, with smaller nests, smaller larvae and poor survival in 1997. Moreover, very few vertebrate prey items were found on both Floney Buzzard nests in 1997, mainly frogs and some juvenile passerines (Table 2). Condition of both surviving Honey Buzzard nestlings varied between nests. The Smilde-bird dropped considerably in mass in the week before fledging, from a maximum of 750 g on 6 August to less than 600 g on 11 and 12 August. This bird also showed retarded growth in wing length. The Boschoordnestling did much better, and fledged at a body mass of >800 g. Both nestlings, however, had many faultbars in their flight feathers (Bij Isma, van Manen & Ottens 1997), The poor condition of the Smildebird led to its death from starvation within a week after fledging, although it still managed to disperse 33 km from the natal site. The Boschoord-bird fledged and dispersed successfully. Overall, the Boschoord-pair fared better under similar adverse conditions as the Smilde-pair experienced, either because their home range included a more varied habitat structure (as it did), or the quality of the pair was better than that of the Smilde-pair, or both (likely). Throughout the 1970s and 1980s, years with low wasp numbers did not lead to large-scale breeding failure in Honey Buzzards as it did in 1997. It is argued that this was caused by the availability of sufficient substitutes, mainly nestlings of Woodpigeons Columba palumbus. As Woodpigeon densities in the eastern Netherlands in the 1970s and early 1980s reached up to 240-1800 pairs/100 ha of coniferous woodland, breeding pairs were semi-colonially clustered in dense stands of fir and spruce, and peak breeding numbers were recorded in July-September (young production highest in August-September), Honey Buzzards easily counteracted low wasp numbers by depredating nestling Woodpigeons. However, both density and breeding season of Woodpiegeons changed dramatically during the 1980s, as cereals were being replaced by green maize. Presently, Woodpigeon numbers in woodland in the eastern Netherlands have dropped by >90%, the breeding peak in late summer has ceased to exist and clustered breeding pairs in dense stands of Norway spruce and Douglas fir has been replaced by widely scattered breeding of solitary pairs. Consequently, Honey Buzzards facing low wasp numbers are unable to substitute wasp shortage by preying on nestling Woodpigeons. In 1997, Honey Buzzards were therefore experiencing extreme food shortage, leading to desertion, high mortality among nestlings and fledglings, poor condition of surviving young, early departure from the breeding grounds and a delay in moult (many birds did not moult at all in the breeding area).