42 ECOLOGY AND LIFE HISTORY OF THE COMMON FROG 



SO that, for example, a tadpole tends to avoid the neighbourhood of 

 larger specimens of its own species. This would seem to be in the best 

 interests of the small tadpoles, and is only a small extension of the 

 basic idea. There is experimental evidence that in other species kept 

 in aquaria, overcrowding is associated with a diminished growth rate, 

 from some kind of mutual interference with feeding (Lynn and 

 Edelmann, 1936, for R. syhatica, Deansley and Parkes, 1945, for 

 Xenopus laevis). There is, of course, no doubt about the importance 

 of this competition. To be late in metamorphosing is to risk desicca- 

 tion or starvation, as the examples at Dagger Lane in 1947, or at Lower 

 Parkfield in the same year, show quite clearly. 



At this point, it may be useful to leave the detail for a time and to 

 examine these size differences from another more general point of view. 



To begin with, how real are these differences? The graphs show the 

 size of the differences in the means, but not their significance. Although 

 at the ponds the phenomenon was striking, it is quite easy in a long 

 series of results to allow the attention to concentrate on the noteworthy 

 events, and to ignore what might be the more usual case. I therefore 

 subjected the whole series of collections at Dagger Lane 1947 to an 

 analysis of variance (Chapter 10, Appendix i a) with a result so highly 

 significant that the further analyses that were planned were obviously 

 superfluous — there is no doubt whatever that this is a general pheno- 

 menon, and a big one. The next question is: are the differences 

 surprising? If tadpole ponds were large places, no one could be 

 surprised to fmd colonies in distant parts of different sizes, but these 

 ponds are so small that a large tadpole could swim across them in a 

 few minutes. The expected event is surely complete randomization: 

 what happens is marked segregation. 



According to my observations, only about half the ponds which 

 contain spawn support tadpoles all through the season as far as meta- 

 morphosis. In the rest, the tadpoles disappear at various dates in the 

 season. It is, of course, impossible in practice to prove that a pond 

 contains literally no tadpoles, but, even if there were a few in the ponds 

 I have regarded as having none, it seems certain that the frog population 

 of a district comes from very few of the ponds in it. 



It may well be felt surprising that tadpoles living at hberty in the 

 relatively vast spaces of a pond, well stocked with animal and vegetable 

 life, should be so short of food. The explanation lies in the remarkable 

 physiology of the animals, which forms the subject of the next chapter. 



