By addition of 1956 observations, this report alters and supplements the resume 4-6 in the Misc. Rep., No. 7, p. 289 as follows:
1. Daily winter feeding range (rice paddies) of the flock roosting at Shinhama is delimited to 12km at the most by hills and scattered villages, while Koshigaya flock in a wide field (30km north of Shinhama) covers 20km, its periferal flocks meeting with those of Shinhama at the east of Tokyo. The feeding density ('lowest density') ranged from 0.4 to 4 birds per 1 ha.
2. Number of breeding pairs is given by each subarea studied, and the factors influencing the breeding density or preference are discussed. Chief factors may be: 1) enough nesting hollows, 2) type and age of vegetation, 3) structure of feeding area (at least some rice paddies) and 4) safety, etc.
3. Rice paddies become unfit for feeding area as the transplanted rice grows. This happens after the young of the 1st brood joined the flock and thus, only few early 2nd broods could be raised (Of the 19 nest-boxes, 17 were used in the 1st and 3 in the 2nd broods; more should be raised in city zone). Then the flock finally moves to the northern field.
4. As season advances, the condition of feeding ground changes by cultivation. Varieties in the insect fauna contained are shown by several sample collections made at different places and dates. Early in season, the ground spiders are dominant everywhere, but the most important food for raising young is the mole-cricket (Gryllotalpa), which is best sought in the paddies ready for rice transplantation in late June, when the nestlings grow and leave the nest. This may be said to support Lack's theory of anticipatory adaptation.
5. Preliminary consideration is given on the method of determining the ecological niche of the starling quantitatively. The quantity of food taken in a day (or an hour) within a unit area by a flock of starling occurring in a study area may be culculated by: L=Nnt/A×W When N: the number of individuals of the flock, n: the times (frequency) of occurrence of the flock in a day (or an hour), t: length of one feeding stay (mean) of the flock in a study area, W: quantity of food taken by one starling within a unit time (1min., 10min., etc.), A: dimension of study area. In breeding season N, n are substituted by the total number of individuals coming in from the colony. Discussion is given of the method of observation and culculation of t and W, and the method to estimate the quantity of food taken by parent birds was suggested from feeding behavior and the time of absence at the nest site.
