PROCEEDINGS OF THE ASSOCIATION FOR THE STUDY OF ANIMAL BEHAVIOUR 605

keep to one row only per trip in orchards of groups, each near the centre of the area it is to dwarf trees. This behaviour may be reflected pollinate. Colonies should not be taken to the in the fruit set of trees needing cross-pollination. orchard until flowering has begun otherwise

Bees tend to forage near their hives, particu- the bees often become conditioned to foraging larly in unfavourable weather, so colonies for on other species which they will not readily orchard pollination should be sited in small forsake when the orchard flowers.

FORAGING AND FEEDING BEHAVIOUR IN THE THREE-SPINED STICKLEBACK. By L. DE RUITER, & J. J. BEUKEMA, The Zoological Laboratory of the University of Groningen, (Haven) the Netherlands.

Observations under standard conditions on single sticklebacks, kept in either small con- tainers with an excess of prey (Tubifex, Enchy- traeus, Drosophila larvae) on the bottom, or in a maze of 18 intercommunicating hexagonal cells with prey scattered at random at constant very low density, lead to the following conclusions :

1. Food searching activities increase with deprivation, decrease with satiation.

2. Searching is not random; the fish learns to explore the hunting area in a systematic manner.

3. The amount of searching at a given ‘hunger’ level depends on the nature of the prey the fish ‘expects’ to find.

4. A complete response to the discovery of prey consists of fixating, grasping, and swallow- ing. Initiated responses may break off at any stage.

5. The probability that encountered prey will release a response increases with deprivation and decreases with satiation, as does the probability that an initiated response will be completed.

6. ‘Reaction distance’ (i.e. distance between fish and prey at the moment of the first visible response) is independent of deprivation (16-88 hours), but increases with satiation.

7. Reaction distance is strongly influenced by learning processes; its end level depends on palatability and conspicuousness of the prey.

8. Relatively unpalatable prey is refused at a much lower level of satiation, if the fish ‘ex- pects’ to find better food elsewhere in the hunting area.

9. The stickleback feeds rapidly until its stomach is full; from then on it feeds at a much lower rate, approximately equalling the rate of digestion.

10. The total amount ingested daily may be independent of the calorific value of the food in the range studied.

11. In fishes starting with an empty stomach, the rate of initial filling up increases with the length of previous deprivation. Thus at least two factors govern feeding in this species: one gastric (‘satiation’) and one systemic (‘de- privation’). In addition temperature, time of day, deprivation schedule, nature and density of food, and learning processes, play important roles.

12. Although practically all aspects of feed- ing and foraging behaviour mentioned above are affected by all these factors, none of the former by itself provides a simple measure of ‘hunger drive’ on which reliable predictions of the other aspects of the behaviour of the stickleback towards its food can be based.

A TYPICAL USE OF THE CILIARY FEEDING MECHANISM IN SOME ANURAN LARVAE. By J. S. KENNY, Zoology Department, Birkbeck College, London.

In typical Anuram larvae the primary feeding the primary mechanism is involved. Instead, mechanism involves the reduction of organic material into particles by means of the buccal

the tip of the tail is modified into a flattened

rasp, the removal of these particles from the filament, which beats independently of the rest

respiratory stream by filtration, the entangle- of the tail, and which is capable of maintaining

ment of the particles in strings of mucus secreted by its slight propulsion the position of the tad-

from special organs, and the transport of mucus pole in mid-water. The diet of this tadpole is

and particles to the oesophagus by ciliary action. mainly phyto-plankton and zoo-plankton. Mid- The typical way in which this mechanism is water feeding is typical in the Pipidae and some used is by the tadpole browsing over encrusting Microphylidae and in both these groups the growth or aquatic plants or other debris. The filters are considerably more extensive and the tadpoles of Phyllomedusa rely for their diet buccal rasp lacking. In some Microphylidae the mainly on particles in suspension in mid-water lips are modified into a shallow up-turned funnel or floating on the surface. No modification of which assists in surface feeding.