Introduction
After advertising in a local bird magazine (De Korhaan) and a regional newspaper (de Gooi en Eemlander) in the early spring of 1997, 18 people applied for participation in a study to count the number of House Sparrows in the vicinity of their homes. The aim of the study was to shed some light on the numbers and distribution of sparrows in Gooi & Vechtstreek, a region of some 25,000 ha. east of Amsterdam, surrounding the city of Hilversum. Half the area is completely urbanized, the other half consists of a mixture of farmland, lakes and woodland. The region is bordered by the Amsterdam-Rhine canal in the West, the Gooimeer and Eemmeer (formerly parts of the IJsselmeer) in the North, the river Eem in the East and the border with the province of Utrecht in the South.

It is considered important to know more about the numbers and distribution of House Sparrows, since there are strong indications that they show a steady decrease in numbers since the early 1980's in several parts of the world. Decreases of upto 50% have been published not only in the Netherlands, but also in the USA, Canada and the UK. It almost looks as if something has changed in the life-style of the people of the western hemisphere which has conflicted with the environments in which House Sparrows thrive! The question that has inspired me was: If this decrease is such a wide-spread, general phenomenon, do we have to worry about the continued existence of the House Sparrow in our own neighbourhoods?

Methods
To study this question I asked the participants to throw some seeds and grain (a commercially available mixture for seed-eating birds) each day at a fixed time and on a fixed spot on the ground in the neighbourhood of their home over a period of 14 days. It is well-known that House Sparrows prefer foraging on the ground and learn very quickly where and when food is available when provided on a regular basis. After approximately one week such a feeding place will be visited by a fixed group of regular feeders, who often gather together in the bushes surrounding the feeding place some time before the scheduled feeding time. Participants were free to choose a suitable, fixed time for feeding between half an hour after sunrise and an hour before sunset. After application of the food they were expected to count how many House Sparrows were present for half an hour. Each day the maximum number was noted down on a form which was returned to me at the end of the 14-day counting period, together with answers to a questionnaire about the location of the house and details about the residential area. There were 2 periods of counting: one in the spring between 1 March and 1 April just before the start of the breeding season, and then one in the autumn between 15 september and 31 October after the end of the breeding season.

Results
The survey results of the first spring (1997) varied considerably. Not a single House Sparrow was observed for the whole period around 5 residences in Hilversum, Bussum (2x), Naarden and Loosdrecht, whilst in contrast there were regular counts of between 25-40 House Sparrows around other houses. When I correlated the results of the counts with the age of the houses, something quite unexpected emerged: low numbers of House Sparrows were observed around older houses, high numbers around younger houses. The dividing line between these two groups was lying at approximately 1950-53!

As it turned out, half of the group of 18 participants owned properties built before 1953 and half built after 1953. When I took the average of the daily maximum numbers of House Sparrows observed during the 14-day counting period, it appeared that an average number of only 3 House Sparrows were attracted to the feeding places close to houses built before 1953, whereas an average of 19 were observed in the surroundings of younger buildings (see Fig. 1).

Such a high statistically significant difference (Mann-Whitney U-test, p<0.01) is supported by the fact that 7 out of the 9 participants who owned a house built before 1953 indicated that they had observed a clear decrease in House Sparrow numbers in their residential areas during the previous 5-10 years, in contrast to 7 out of the 9 participants with younger properties who claimed that they had noticed no change in numbers during the same period.

Discussion
What causes these differences is a complete enigma to me. I can imagine that in older residential quarters not only the houses are older but also the private and public gardens and parcs. It is quite conceivable that through the eyes of a House Sparrow the vegetation of these areas has gradually become too mature and has deteriorated in other ways resulting in a reduction in the food resource and foraging potential for House Sparrows. It clearly calls for further investigation.

However, solving the enigma will not be simple. Further study should set out to explain why in one area mortality and eventually emigration exceeds the yearly increase of young birds and eventually immigration, resulting in low population levels, whereas in another area the situation is reversed, resulting in stable, high population levels. This is a complex matter.

It will be necessary to infer reliable estimates of death causes for different age classes, and to collect breeding success data over a number of years. This can be achieved by trapping and marking individuals with colour rings. At the same time one should also try to find out where the surviving young birds will settle for their first breeding season. The study ideally should be supported by an analysis of different food items and an assessment of the quantity of food that a House Sparrow eats throughout the year. This presents a long challenging action plan.

Therefore it seems likely that the definitive solution to the problem of the decline of the House Sparrow will not come easily. Yet, in the interim, there are some conclusions that can be drawn from this study. Many people consider Magpies and to a lesser extent also Carrion Crows, Jackdaws and Jays as the main predators, since these birds may plunder the nests of small songbirds, particularly when they have to care for their own young. However, the role played by these predators is necessarily rather limited, since there are no differences in the numbers of the most common ones (Magpies and Jackdaws) in residential areas with houses before and after 1953 and therefore the potential to become a victim of a Magpie will be approximately equal in either area.

Apart from that nobody knows the extent of Corvid predation on House Sparrows, as sparrow nests are generally inaccessible to these birds and within a few days of fledging young sparrows are perfectly capable of avoiding contact with Magpies and Jackdaws, although they are most vulnerable in the days when they first leave the nest.

As mentioned the risk is greatest during the Corvid breeding season, which is much shorter than that of House Sparrows. House Sparrows produce 2-4 broods each year and their breeding season extends from May to September. Later broods of the House Sparrow will suffer little or nothing from Magpies, perhaps sufficient reason not to proceed to an immediate widespread persecution of Magpies, Carrion Crows and Jays.

It would be unwise to draw wider conclusions from the present data as the sample is limited and further analysis couls lead to mere speculation. It is my intention to recruit more participants and with greater participation it might be possible to construct a map showing a division of sparrow-rich and sparrow-poor areas. Such a map could then be compared with a habitat map of the region and lead to a greater understanding of the causal factors involved in the decline of the House Sparrow.